1 Wednesday, 2nd November 2022 2 (10.00 am) 3 MR STUART SWEETMAN (continued) 4 Further questioned by MR STEVENS 5 MR STEVENS: Good morning, sir, can you see and hear me? 6 SIR WYN WILLIAMS: I can certainly hear you and no doubt 7 I will see you in a moment. 8 MR STEVENS: Mr Sweetman, can you see and hear me? 9 A. I can hear you but not see you. I can see the Chairman. 10 MR STEVENS: I wonder if that may be fixed. If you bear 11 with us a moment. 12 (Pause) 13 Sir, I'm told five minutes will be required to 14 address that so you can see me. 15 SIR WYN WILLIAMS: All right. Then we will just break off. 16 I will come back visibly in five minutes, all right? 17 MR STEVENS: Thank you, sir. 18 (The Inquiry was paused for technical reasons) 19 MR STEVENS: Sir, can you see and hear me? 20 SIR WYN WILLIAMS: Yes, I can and I can see Mr Sweetman and 21 I guess he can see me? 22 A. I can see the Chairman and I can see you, sir, yes. 23 I can see both of you. 24 MR STEVENS: Excellent. 25 Well, Mr Sweetman, you were giving evidence 1 1 yesterday and you are going to continue to give that 2 now. I will pick up where we left off. We were 3 discussing, towards the end of the day, Acceptance 4 Incidents and the workshops that went into it and you 5 were briefed on the progress of those acceptance 6 workshops as they were progressing; is that correct? 7 A. Probably, I think the record shows that. I don't recall 8 that. 9 Q. We were yesterday discussing one of the Acceptance 10 Incidents, which is number 376, concerning data 11 integrity and discrepancies in the cash account, yes? 12 A. Yes, we were. 13 Q. Now, eventually, in the second supplemental agreement, 14 Post Office and ICL Pathway agreed in respect of that 15 incident that between 3 October 1999 and 16 14 January 2000, the percentage of cash accounts 17 received across the TIP interface, with discrepancies, 18 should not exceed 0.6 per cent. Do you recall that? 19 A. I don't. 20 Q. Yesterday, we discussed that you looked at the bigger 21 picture when considering whether an error rate was 22 material. 23 A. Yes. 24 Q. Did you listen to the evidence of Mr Copping earlier 25 this week? 2 1 A. I haven't. 2 Q. This was put to him, this agreed rate of 0.6 per cent 3 for the rate of discrepancies, and it was put to him 4 that when applied across the entire Post Office network, 5 that would amount to 100 cash account discrepancies in 6 any week, which he considered to be quite a significant 7 error rate. Would you agree with that? 8 A. I suppose I would have to compare it with the 9 pre-Horizon error rate that arose within the cash 10 accounting system, the old manual system because there 11 were -- I don't know the number. But there were lots of 12 errors coming out of offices which had to be corrected 13 and there was a team of people in Chesterfield who 14 picked up those errors, analysed them and did what was, 15 therefore, necessary to correct the records. That 16 sometimes was adjustments within head office, sometimes 17 it was going back to the office and saying, "Can you do 18 this on the next cash accounts and that will bring it 19 back into balance?" 20 I think there were dozens of people in Chesterfield 21 doing that correction process and they would have the 22 capability of handling that level of errors because it 23 was common practice. So, that being in place would have 24 given me personal comfort that that low level of error 25 rate that we had defined in that agreement you quoted 3 1 was handleable and I think that's what I would have 2 assessed the issue as. 3 Q. So, before moving on, again on this point of 4 materiality, I think what you are saying is, from your 5 perspective, this error rate passed the materiality 6 threshold which you were applying and you would be happy 7 with that level of errors, yes? 8 A. You are never happy with any errors but if you can 9 handle them, that it doesn't undermine the 10 completeness -- the truth and fairness of the record 11 keeping, then, from my perspective as a director, 12 I would be happy that we are running an accounting 13 system under control because we can pick up the errors, 14 we can correct them and move forward. 15 I was looking at it from an organisational basis, 16 not a subpostmaster basis. 17 Q. Yes. You said yesterday that the perspective from 18 a subpostmaster within -- who is affected by one of 19 these errors would be different. Was anyone within Post 20 Office Counters representing the interests of 21 subpostmasters at this stage? 22 A. Well, the whole management structure. From what I have 23 read and I think I remember, we had an organisational 24 structure where maybe 50 post offices were looked after 25 by a retail network manager and during this 4 1 implementation phase we had specific managers with that 2 sort of line experience looking -- holding the hands of 3 and looking after the subpostmasters who were going 4 through the implementation. They would have been there, 5 on the spot, seeing what was coming back and what was 6 causing errors in the local office and helping the 7 subpostmaster respond. 8 Q. I think we are talking at cross-purposes? 9 A. Are we? Sorry. 10 Q. I'm discussing the considerations that were taken into 11 account for whether this 0.6 per cent figure was 12 material and in your evidence yesterday you compared the 13 different perspectives. On the one hand, from your 14 perspective, looking at it from Post Office Counters as 15 a whole and then on the other hand from an individual 16 subpostmaster who would be affected by an error. 17 Now, when considering this level of errors, I say in 18 that context, was anyone at Post Office Counters 19 representing or considering the interests of the 20 individual subpostmasters in respect of that error rate? 21 A. The honest answer is I can't remember. But I suppose 22 that was down to all of us. But no one specific had 23 that hat on to say "In this meeting I'm representing the 24 subpostmaster". I don't think that was in place but 25 I really do not remember, sorry. 5 1 Q. I think you may have answered this already but just so 2 the question is put to you, are you aware of anyone 3 seeking the views of subpostmasters on that issue? 4 A. We had lots of feedback routes from subpostmasters. At 5 national level: the general secretary, the president. 6 At regional level there would be feedback mechanisms and 7 there would be mechanisms down in subpostmasters and in 8 the regions. So there would have been various lines of 9 communication. What those specifically were about this 10 error type, I don't recall, sorry. 11 Q. I want to jump now to January 2000. Please can I turn 12 up POL00000336. 13 This is a meeting on the Post Office board held on 14 11 January 2000 in which you attended. Please can we 15 turn to page 11. 16 Can we go further down, please. Sorry, slightly 17 further up. Actually, I'm so sorry, if we start on the 18 page before that at the bottom. 19 Sorry for the delay there, Mr Sweetman. 20 A. That's okay. 21 Q. This concerns an update on the Horizon programme and it 22 refers to the rollout of Horizon being due to recommence 23 and that a great deal of work had been undertaken to 24 rectify difficulties identified in three areas: 25 "Systems stability; 6 1 "Accounting integrity and; 2 "The provision of support to officers 3 "Although as yet uncertain, it was anticipated that 4 these issues would not prevent rollout recommencing." 5 Go over to the next page slightly please. 6 "Given that the programme was expected to recommence 7 rollout, it would be helpful for the Board to understand 8 what marketing opportunities were now being considered." 9 A. Yes. 10 Q. I understand you don't have any recollection of this 11 meeting? 12 A. Not that specific meeting no but I can read the words 13 and I can interpret them. 14 Q. So at this stage, were the board asking you any 15 directive questions about the accounting integrity issue 16 which we have been discussing? 17 A. The minutes record -- so this board meeting would have 18 been earlier in the month prior to the recommencement of 19 the rollout and the board would have had a report, which 20 I'm not sure if I have seen, which says "We are working 21 on these issues and if those issues are satisfactorily 22 resolved we will recommence rollout". 23 I don't recall whether -- how far I was pushed on 24 that or asked about it -- but they could well have 25 asked, but I don't know. 7 1 Q. Please could we turn to FUJ00118186. This is the third 2 supplemental agreement between Pathway and Post Office 3 Counters dated 19 January 2000, just over a week after 4 that board meeting. So presumably this was being 5 discussed around that time of that board meeting we were 6 discussing? 7 A. Yes, it must have been. 8 Q. Can we go to page 2, please, and to (d), thank you. 9 It says: 10 "By clause 6.1.1 of the Second Supplemental 11 Agreement, POCL has a right to postpone the resumption 12 of rollout from January 2000 if any of the criteria in 13 parts a to c of Schedule 4 to the Second Supplemental 14 Agreement shall not have been met by 24 November 1999." 15 (e): 16 "Both POCL and the Contractor acknowledge that at 17 least one of those criteria were not met and accordingly 18 that the right contained in clause 6.1.1 became 19 exercisable." 20 Then please can we turn to page 5, paragraph 5.3. 21 We're back to the data integrity issue and it says that: 22 "The Contractor shall from the date of this 23 agreement until the end of the TIP integrity Checking 24 Period make available to POCL promptly upon request 25 appropriate experts to explain to POCL the Contractor's 8 1 analysis of all root causes of Cash Account 2 Discrepancies and the measures which the Contractor 3 shall have implemented in order to prevent the 4 recurrence of any Cash Account Discrepancies which would 5 not have been detected by the Accounting Integrity 6 Control Release." 7 Do you accept at this stage cash account 8 discrepancies remained an ongoing problem? 9 A. I think that's evidence that they were. 10 Q. By this clause it was envisaged that the reconciliation 11 processes that were supposed to identify cash account 12 discrepancies may fail to do so? 13 A. I could read that into those words. They are quite 14 technical words. The end bit, which says "which would 15 not have been detected by accounting integrity control", 16 does imply that, yes. 17 Q. Would you have raised these concerns with the Post 18 Office board at the time? 19 A. Not at this level of detail, I would not have thought. 20 Apart from the generalities which were noted by the 21 board, that the errors were still being worked on and 22 improvements sought. So, in general terms, I think what 23 the board minutes record is us saying "There are still 24 issues that we are managing". This is a level of detail 25 that we would not have, in those words, taken to the 9 1 board. 2 Q. Let's move to the March board meeting. That's 3 POL00021469. The 14 March 2000 board minutes, again at 4 which you attended. Please could you turn to page 7 of 5 this. 6 If we could move down, sorry. Thank you. This is 7 described as "Commercial Development of the Horizon 8 Platform in Post Office Network". It refers to, 9 firstly, the rollout and then it moves on to discuss the 10 other commercial uses of the Horizon platform? 11 A. Yes. 12 Q. At this stage, this is following the third supplemental 13 agreement. Did you discuss with the board in any detail 14 the ongoing issues with data integrity? 15 A. I don't know but, looking at the record here, the focus 16 was on the future, not the current, because the rollout 17 had been accelerating and we had learned all the lessons 18 of the early offices and had put those in place and the 19 rollout was now going at a rate that we had originally 20 planned and so the board would have taken comfort from 21 that but I can't see any evidence that we brought to 22 them anything else. 23 Q. Do you recall whether either John Roberts or anyone on 24 the board asked you any questions about that issue at 25 the time? 10 1 A. No, I don't. 2 MR STEVENS: Thank you, sir. That concludes my questions. 3 But we have some questions from recognised legal 4 representatives. 5 SIR WYN WILLIAMS: All right. Well, my understanding from 6 last night was that Mr Henry and either Mr Stein or 7 Mr Jacob had comparatively few questions. Is that still 8 the position? If so, then they can go first and then 9 Hudgells can mop up, so to speak. 10 MR STEVENS: My understanding is no questions from Howe+Co. 11 MR JACOBS: That's correct, yes. 12 MR HENRY: Sir, no questions on behalf of Hodge Jones & 13 Allen either. 14 SIR WYN WILLIAMS: All right then, straight over to -- is it 15 Ms Patrick or Mr Maloney who is going to ask the 16 questions? 17 MS PATRICK: Sir, it is me, Ms Patrick. 18 SIR WYN WILLIAMS: Fine. 19 Questioned by MS PATRICK 20 MS PATRICK: Good morning, Mr Sweetman, can you hear me? 21 A. Good morning, Mr Patrick. Yes, I can. 22 Q. And can you see me? 23 A. I can. 24 Q. Thank you. We only have a very few questions and it is 25 on two topics. The first question we have is about one 11 1 document and it is at POL00031230. 2 Is that now on the screen for you? 3 A. Yes, it is. 4 Q. This was an internal review within POCL, commissioned 5 from the POCL finance director in January 1999 and, at 6 that time, you were the head of the POCL management 7 team, weren't you? 8 A. Yes, I was. 9 Q. It is not marked as to who received the document but, as 10 the MD, would it have been provided to you? 11 A. Yes, it would have been. 12 Q. Thank you. If we can go to page 2, and we are looking 13 at 2.4, please, which is the second paragraph from the 14 top. I will read it out, so that we have all got the 15 same understanding: 16 "The conclusion of negotiations with a firm decision 17 to proceed should put an end to a protracted period of 18 uncertainty, permitting a fresh start with renewed focus 19 not only for the Horizon project but for the POCL 20 business." 21 It is this next part I want to look at particularly: 22 "Unfortunately, many uncertainties, unanswered 23 questions and doubts about the future remain, so that 24 the benefits of such a fresh start seem unlikely to be 25 obtained without a concerted, focused and single minded 12 1 leadership effort by both POCL and ICL teams to 2 emphasise the positive." 3 If we can skip down a little to paragraph 2.6. 4 I think you can see that there on the screen 5 Mr Sweetman? Can you see paragraph 2.6? 6 A. Yes, I can. 7 Q. That starts: 8 "Several senior managers, close to the project, but 9 not principal negotiators, whose judgement I respect, 10 express significant reservations about the risks of 11 proceeding. These centre on their continuing doubt 12 about the ability of ICL to deliver a satisfactory 13 product; the absence of transparency in the PFI 14 contract; the risk that ICL's financial fragility will 15 endure throughout the project, with the possibility of 16 repeated claims on The Post Office for extra 17 contributions (which, by then having no alternative, it 18 will be unable to resist); and doubts about POCL's own 19 ability to give it the focus essential for success. 20 Observation of the track record so far offers reasonable 21 foundation for such views." 22 If we could skip over to page 6 of this document, 23 now I also want to look at paragraph 2.24, which is 24 a little bit further down on that document as you can 25 see it now. 13 1 This is the summary of his findings: 2 "In summary, there are drawbacks and uncertainties 3 with going ahead, but they are not greater than those 4 associated with termination. Going ahead will require 5 very heavy, single minded commitment to Horizon and to 6 the partnership with ICL in order to minimise the 7 drawbacks." 8 If we skip to the very next paragraph, 3.1, which 9 I think you can see. It goes on with his 10 recommendations: 11 "The high-profile of the Horizon renegotiation, and 12 The Post Office's support of Horizon against significant 13 opposition mean that failure to make the deal stick with 14 a successful implementation would be politically and 15 commercially extremely damaging." 16 This part is in bold: 17 "It is of great importance for the credibility of 18 The Post Office (not just POCL) that it should be seen 19 to have judged the debate correctly and made the right 20 decision." 21 If we skip further down to 3.2, which goes over the 22 page, his recommendations continue: 23 "Furthermore, POCL's commercial success will now 24 depend heavily on Horizon. It will not have the funds 25 for alternatives. Horizon must therefore be implemented 14 1 in a way that which ensures achievement of service and 2 commercial goals, for customers and clients. To be sure 3 of this, the approach to developing the business vision 4 will need to be adapted to become visibly 5 Horizon-centric. People's enthusiasm for finding 6 workarounds and alternative approaches will need to be 7 firmly channelled into making Horizon deliver what is 8 required. It is likely that this Horizon-centricity 9 will need to apply beyond the bounds of the current POCL 10 business. Shaping for Competitive Success will need to 11 ensure that organisation boundaries facilitate effective 12 operation of Horizon and the ICL partnership, and not 13 make it gratuitously more difficult." 14 Now that we have read that and we have all got the 15 same understanding, yesterday I think you said POCL had 16 to be sustainable in the post-ACT world and I think you 17 agree that to do that, first, Horizon had to work; is 18 that right? 19 A. Correct. 20 Q. I think you said yesterday Ian McCartney may have told 21 you, or something like, "I'm going to make sure everyone 22 knows in the Post Office that this has got to work"; is 23 that correct? 24 A. Yes, yes. 25 Q. So POCL would have had external pressure to adapt to 15 1 become visibly Horizon-centric; is that fair? 2 A. But we wanted -- yes, we wanted to be Horizon-centric, 3 that was our strategy. 4 Q. And what I'm saying is you wanted that -- 5 A. Yes. 6 Q. -- but there was also external pressure to achieve that 7 approach? 8 A. Isn't it an iterative process? We wanted Horizon 9 because what it would equip the company to do in the 10 marketplace and we went out and sold the idea to the 11 politicians and the senior management and the board that 12 if we had this computerised network -- modern network, 13 we could do these things, we could actually sustain our 14 commercial future. So we were broadcasting "That's what 15 we want" and that bounced back to us saying, "Yes, okay, 16 you get on with it but make sure you deliver on the 17 targets", and that's the management process. So I don't 18 think they were over-the-top pressures, I think that 19 people were playing their roles. Politicians were 20 making big decisions and they had very difficult 21 decisions to make, the board were taking those on board, 22 the chairman and chief executive had that obligation to 23 deliver to the political needs, the shareholders' needs. 24 That then got disseminated down to the businesses and we 25 all responded. 16 1 This particular report, I think, was by Roger Tabor, 2 who was the then finance director of Post Office network 3 or Post Office Limited and he was a very professional, 4 analytical individual and I think he would have been 5 asked to produce this report to make sure, as a team, we 6 had everything -- do we have all this in place? Do we 7 have our own confidence that we have this in place to 8 deliver? And this is an internal document which says 9 "Are we fit for purpose"? And he was telling us, "All 10 these things need to be in place if you are going to 11 deliver what the business plans are". 12 Q. Mr Sweetman, thank you. So it was becoming visibly 13 Horizon-centric, the message which you and you staff and 14 your board, your team, adopted from then on? 15 A. It wasn't from then on. Running up to it -- I think 16 Roger, in this report, was saying, "You have got to 17 continue it. Don't back off from this core strategy 18 that you are following of equipping the network of 19 postmasters with a modern computer system. You can't 20 back off of it because there isn't a plan B". 21 Q. Thank you. Sir, I'm going to move onto the second topic 22 that we wanted to ask about. You said yesterday, in 23 detailed discussions, the board would have expressed 24 concerns about outstanding technical issues. I want to 25 look at two later board meeting minutes and the first is 17 1 a meeting minute from May 2001, which is RMG00000009. 2 We have got it on screen and I think what I would like 3 to do is -- On the first page we can see that you were 4 present and if we can scroll down to page 5, the part 5 that I want to deal with is on page 5. That was the 6 1 May 2001. 7 Can you see that document, Mr Sweetman? 8 A. Yes, I can. 9 Q. We can see at the top there, there is a heading "error 10 reconciliation and accounting" and I think a paper 11 number and "Horizon beyond 2005" and again a paper 12 number. I'm going to read underneath so we are all 13 looking at the same document: 14 "The Board considered the two papers from Stuart 15 Sweetman, and the accompanying presentation by David 16 Smith, which gave an illustration of the fundamental 17 nature of the change to the operating process that would 18 be delivered by the Error Reconciliation and Accounting 19 Project (ERA). ERA needed to be viewed in conjunction 20 with options for the future of Horizon beyond 2005 which 21 included inter alia an approach to securing internal 22 funding for ERA." 23 It goes on to say things that the board noted in 24 relation to that project, or that plan and if we could 25 scroll down to (d). I think you can see that, 18 1 Mr Sweetman? 2 A. Yes, I can. 3 Q. "Success of the ERA project would depend on client 4 cooperation in redesigning their own forms and 5 procedures. Most current clients were keen to be 6 involved, especially now that they were able to see 7 Horizon as a working reality." 8 If we could scroll down to (e). I think we are 9 going to have to go further. If I could bring you up to 10 the top of the page that would be helpful. It goes on: 11 "the success of the Horizon installation programme 12 had helped increase confidence in ICL as a supplier. 13 However they were not necessarily the best future 14 supplier, but the recommendation was to continue with 15 them as the potential disruption to the PO network from 16 starting a new procurement process to source a new 17 supplier now would be considerable, and require 18 a rewriting of the strategic plan. However it was 19 recognised from a negotiating point of view it was 20 highly risky to give ICL a commitment for the future, 21 and more thought should be given to our negotiating 22 stance ..." 23 It goes on a little to say some more about: 24 "It should not be assumed automatically that ICL's 25 funding of the investment was the only option. It may 19 1 be preferrable ... to seek external funds ..." 2 If we could go over the page please it goes on to 3 talk about what the executive board agreed and it says: 4 "while, inevitably, there were a large number of 5 unresolved issues about ERA and Horizon, their general 6 direction was right and the respective strategies should 7 be worked up further in the light of this discussion, 8 and would be taken to the July Consignia Board as 9 a strategic issue." 10 That last point, what did you understand to be the 11 large number of unresolved issues about ERA and Horizon? 12 A. I don't recall a lot of detail to be honest. What 13 I have read there, in my earlier evidence I said that 14 there was a group of people in Chesterfield whose job it 15 was to pick up errors and then chase them down until 16 they have been corrected, either in the centre or in the 17 cash accounting system of subpostmasters. And that was 18 a very big and expensive operation because it was there 19 and, in a perfect world where systems are working 20 absolutely perfectly, it is a wasted cost, a big lump 21 of -- a big group of people correcting errors. 22 If errors aren't created, you don't need those 23 people. And I think ERA was about smartening up that 24 whole process of picking up and correcting errors. 25 That would need to take input from Horizon but would 20 1 be a sort of central accounting reconciliation system. 2 I'm sort of thinking through what this would have been. 3 And, clearly, how you link that central reconciliation 4 process to the detail of Horizon in local offices would 5 take some designing and some development and that is 6 what I think this refers to. But I can't absolutely 7 remember. This is me putting words -- 8 Q. Mr Sweetman, I may be able to assist you. In the 9 interests of speed, there is a part of that minute that 10 I did not read and I am sure you can see it at the top 11 of the page there: 12 "ERA would cause at least 400 job reductions in 13 Chesterfield, and a plan for managing these -- and the 14 associated negative PR, would be needed." 15 What you have said about error corrections happening 16 at Chesterfield and that team being there, does that 17 note there help your memory? 18 A. I hadn't remembered it was of the order of 400 but, 19 certainly, to keep on top of the error rates and 20 correcting them to make sure we were keeping proper 21 books of account, we had invested heavily in this 22 correction process. Now, if we could develop a system 23 which did that electronically, computer to computer, we 24 would then, you know, coldly, be able to save the cost 25 of 400 jobs. 21 1 Now, I think there were 2,000 or 3,000 people 2 employed in Chesterfield. We were the major employer in 3 Chesterfield and we would have to handle that not only 4 negative PR but, from an industrial relations point of 5 view, that would all be very sensitive issues to manage. 6 Q. Thank you, Mr Sweetman. So this may be a project, from 7 your recall, that would be looking at adjusting how 8 errors are handled once they are being managed. Looking 9 at the record, there is no record here of the board 10 asking for any further information or any update on what 11 the management of reconciliation or balancing errors, 12 which were happening in Horizon, looked like, is there? 13 A. There isn't. 14 Q. It says of course, we have gone to that note: 15 "... inevitably, there were a large number of 16 unresolved issues about ERA and Horizon ..." 17 But there is no note here or record of the board 18 asking for any further information on what those 19 inevitable and seemingly unresolved issues were or were 20 likely to be, is there? 21 A. No, there isn't. 22 Q. In hindsight, was the board here being invited to 23 approve a commercial project -- 24 A. No. 25 Q. -- without sufficient information about underlying 22 1 issues which might relate to their decision? 2 A. No, they weren't authorising it, no. If you page back 3 up, it will give the context to which all this was being 4 discussed, I think. 5 Q. I think if we go back to the page we were on, I may be 6 able to assist, Mr Sweetman. If we go back to point (e) 7 and down and over the page, you will see there the board 8 was saying their general direction was right and the 9 respective strategies should be worked up further -- 10 A. Yes, I think what -- to understand the management 11 process within the Post Office, I was bringing this to 12 my colleagues on the executive board as, "In our plan, 13 we are thinking of -- we need to spend this sort of 14 money on doing this sort of thing. Can we have support 15 for that direction?" This is what this discussion was 16 about. "Is it the right direction to go?" They were 17 not all -- they were saying "Yes, go away and work out 18 the detail and come back with a project to be 19 authorised", and then that would have been authorised by 20 the Major Projects Committee and, if it was big enough, 21 by the board. This was a nod through, "This is the 22 right way to go, go ahead and develop a project". 23 Q. A nod through but at no point, or no record there, no 24 record there of them asking about an update on 25 reconciliation, or balancing errors, or for any further 23 1 information on what the inevitable and unresolved issues 2 were likely to be at that point. 3 A. No, there was no record there and it doesn't surprise me 4 that there was no record because that was not the 5 purpose of the discussion. It might well have come out 6 of the discussion but it was not the purpose of this 7 discussion. 8 Q. Thank you. I'm going to move on to the last document 9 that we want to look at, Mr Sweetman, and it is 10 POL00021476. I think we can see you on the first page 11 there and you are in attendance. The date is 12 12 June 2001, so the following month, and we can see 13 there you are now, by this point, group managing 14 director, customer and banking services? 15 A. Correct. 16 Q. Is that a change in role for you? 17 A. Yes. We went through a reorganisation and, in that 18 role, I had the managing director of the Post Office 19 Network reporting to me, so that individual would have 20 looked after, effectively, the operations of the old 21 Post Office Counters network, but I also had strategic 22 directors developing the banking services and also 23 customer services down in Swindon which ran the BBC 24 licensing system. 25 Q. So if we can look over to page 4 and there is only one 24 1 point that's relevant to the Inquiry on that page. You 2 will see, a third of the way down, this is in the chief 3 executive's report which is made clear on the page 4 before. It reads: 5 "Horizon: the Board also expressed its 6 congratulations and thanks to the team working on the 7 Horizon programme, on the successful completion of the 8 installation of over 40,000 machines and training of 9 over 60,000 people in Post Office Network." 10 A. Yes. 11 Q. The thanks are minuted. There is no recollection -- or 12 do you have any recollection of any update then on any 13 impact across the network of the installation, or any 14 record of any request or any update as to the impact of 15 any training as part of that process? 16 A. No, I don't recall that sort of discussion. 17 Q. There is no record there or any update on whether there 18 were, at that time, any recorded problems arising from 19 subpostmasters. 20 A. No, there isn't. 21 Q. And there is no record of any discussion here, is there, 22 it's only a note of congratulations? 23 A. Correct. 24 Q. Would it be consistent with your recall that, after the 25 resumption of the rollout in January 2000, the board and 25 1 the board meetings focused only on the commercial 2 opportunities for Horizon and the successes of Horizon? 3 A. I think that's -- whether it was only, I don't recall, 4 but, principally, the focus was on the future and 5 developing profitable services to operate through our 6 network. 7 MS PATRICK: Thank you, Mr Sweetman. We have no further 8 questions for you. 9 A. Thank you. 10 MR WHITTAM: Sir, Richard Whittam, we indicated yesterday we 11 had very one short point for clarification. 12 SIR WYN WILLIAMS: Fine. 13 Questioned by MR WHITTAM 14 MR WHITTAM: Mr Sweetman, my name is Richard Whittam and we 15 represent Fujitsu. I want to take you back, please, to 16 late August/early September 1999 and it was the 17 Ernst & Young letter we looked at yesterday. I will ask 18 that to come up. POL00090839. On that page, you told 19 us yesterday, Jeff Triggs was the very clever lawyer at 20 Slaughter and May. 21 A. Yes. 22 Q. Can you recall now what role Keith Baines had at that 23 time in Post Office Counters Limited? 24 A. Well, it says there "Horizon Commercial". I wouldn't 25 like to produce a job description now, I really don't 26 1 remember but he was an individual who was good at 2 handling detail and would have ensured that what we 3 agreed with Pathway was reflected in agreements, were 4 properly priced, and that sort of thing. But I really 5 have not a very clear view of that. 6 Q. Don't worry, I appreciate I'm asking you about a single 7 document 23 years ago. But if we could go to the next 8 page, please, just so you can recall its context. The 9 very first paragraph, it is obviously a letter that had 10 been requested by Post Office Counters Limited? 11 A. Yes, that is right. 12 Q. You will recall that David Miller had suggested you 13 might have been the person who requested it and you said 14 you weren't surprised, it could have been you, you 15 couldn't recall at this remove of time. But you did 16 remember -- 17 A. I don't recall specifically requesting it, but it was 18 the sort of thing I might have suggested to him would 19 have been useful in our discussions with Pathway. 20 Q. You described the discussions at this time as the very 21 hard-nosed discussions with Pathway and this would be 22 useful to give you a bit of clout? 23 A. I think that's fair, yes. And I think that's totally 24 reasonable. 25 Q. So this letter, could it be summarised, was requested as 27 1 a tool to be used by Post Office Counters Limited in its 2 negotiations with ICL? 3 A. Not solely as a tool because we wanted to keep the 4 auditors on side. They needed to be able to design 5 their audit processes, et cetera, and it was important 6 for them to understand the details of the system and it 7 was a way of bringing them in as well. 8 I would say giving us a bit of clout was one of the 9 objectives but not necessarily the sole objective 10 because we wanted to know their professional view, 11 because that gave assurance to what we were trying to do 12 on our internal agenda. 13 MR WHITTAM: Mr Sweetman, thank you very much. 14 Sir, that's all that I ask. 15 SIR WYN WILLIAMS: Thank you. I think that now concludes 16 your evidence, Mr Sweetman. 17 A. Thank you. 18 SIR WYN WILLIAMS: I would like to thank you for making your 19 written statement and for devoting the time to give oral 20 evidence to us. 21 A. Would I be permitted to read a final statement that 22 I have written into the record? 23 SIR WYN WILLIAMS: Yes. I had forgotten you mentioned that 24 yesterday, but please do. 25 A. Okay, thank you. 28 1 I have learned from the media and my involvement in 2 this Inquiry of the devastating impact that flawed 3 prosecutions have had on the lives of subpostmasters and 4 their families. They all have my sincere sympathy. 5 These are easy words for me to say but I really do mean 6 them. 7 I am sure that this Inquiry will identify numerous 8 contributing factors of commission and omission that led 9 to what happened. My detailed memory of events of 20 to 10 25 years ago is poor but it has been stimulated by the 11 documents you have questioned me about. I hope it has 12 been shown that throughout my involvement with Horizon, 13 I was, at all times, acting to ensure the long-term 14 viability and sustainability of Post Office Counters and 15 its network of Post Offices. If Horizon had failed, the 16 many thousands of postmasters and Crown Office staff 17 would have lost their livelihoods and communities 18 throughout the country would have been deprived of 19 access to important services that the network provided. 20 I believe I led Post Office Counters by balancing 21 the needs of our stakeholders, our public customers, our 22 paying clients, our employees, our subpostmasters and 23 our shareholder. I was personally driven by the concept 24 of continuous improvement with an equality management 25 framework we call "Customer First". My personal 29 1 integrity has its origins in training and practising as 2 a chartered accountant. I believe that I adhered to 3 those principles throughout the events being 4 investigated. 5 I believe I had well-founded confidence in the 6 people and teams that reported and supported me in my 7 role. When I used to go into my local Post Office, it 8 was with a sense of pride to see the Horizon terminals 9 being used. That pride has now been tarnished by the 10 revelations that this Inquiry is investigating. For the 11 record, I don't believe I have had any personal 12 involvement in the prosecution decisions that this 13 Inquiry is investigating. I'm appalled by what has been 14 revealed about them. 15 Thank you. 16 SIR WYN WILLIAMS: Thank you, Mr Sweetman. 17 MR STEVENS: Thank you, sir. 18 If we may take a short break while we change 19 witnesses here. 20 SIR WYN WILLIAMS: Yes, of course. Five minutes or do you 21 want longer? 22 MR STEVENS: I think 10, sir, that would be appreciated. 23 SIR WYN WILLIAMS: All right a 10-minute break then. 24 MR STEVENS: Thank you, sir. 25 (10.55 am) 30 1 (A short break) 2 (11.03 am) 3 MR BEER: I call Jeremy Folkes, please. 4 MR JEREMY PETER FOLKES (sworn) 5 Questioned by MR BEER 6 MR BEER: Please take a seat, Mr Folkes. As you know my 7 name is Jason Beer and I ask questions on behalf of the 8 Inquiry. Can you tell us your full name please? 9 A. Jeremy Peter Folkes. 10 Q. Thank you very much for giving evidence to the Inquiry 11 and thank you very much for the very detailed statement 12 that you previously provided to the Inquiry. We are 13 very grateful to you for the assistance that you have 14 given and which you will give today. 15 You should have in front of you a hard copy of your 16 witness statement which, excluding the index to the 17 exhibits to it, is 70 pages in length and is dated 18 7 September 2022. Is that in front of you? 19 A. Yes. 20 Q. Could you turn to page 70 please, the last page. Is 21 that your signature? 22 A. It is. 23 Q. Are the contents of the statement true to the best of 24 your knowledge and belief? 25 A. There is one very minor correction on page 39. 31 1 Q. I wonder whether we can go to that. That is 2 WITN05970100 at page 39. 3 A. It is paragraph 114. 4 Q. So page 39. 5 A. It is just -- it says there that Bird & Bird were POCL's 6 legal firm. In re-reading documentation, I realise they 7 were the programme -- the PDA's legal firm, rather than 8 being POCL's. 9 Q. Thank you very much. With that correction are the 10 contents of the statement true to the best of your 11 knowledge and belief? 12 A. They are. 13 Q. Thank you very much. A copy of that statement will be 14 uploaded to the Inquiry's website. I'm not going to ask 15 you questions about every aspect of it, just related 16 parts of it, do you understand? 17 Your background and experience, please. I think you 18 were employed by the Post Office between 1987 and early 19 February 2000; is that right? 20 A. That is correct yes. 21 Q. Working primarily in projects for Post Office Counters 22 Limited? 23 A. Yes. I was working, actually employed by the Post 24 Office IT department but almost everything I did was for 25 POCL projects. 32 1 Q. So far as concerns this Inquiry, your most relevant 2 work, is this right, was on the Horizon project or 3 Horizon programme as it became known, from 1994 until 4 early February 2000? 5 A. Yes. 6 Q. In terms of your qualifications and career before 7 Horizon and indeed after it, I think you qualified with 8 a degree in mathematics, is that right? 9 A. Yes. 10 Q. On graduation you worked for British Gas, then Logica, 11 which I think was a IT and management consultancy firm? 12 A. To be pedantic, I actually worked for British Gas before 13 going to university, I had a gap year, and during my 14 time at university. I did work in IT during that time. 15 Q. Then after graduation to Logica is that right? 16 A. Yes. 17 Q. Am I right; it was an IT and management consultancy 18 firm? 19 A. It was but my work was on IT projects. 20 Q. And you worked on software projects, always holding 21 a technical role; is that right? 22 A. Yes. 23 Q. Whether as a developer, a designer or a team leader? 24 A. Yes. 25 Q. You joined, as I have said, the Post Office in 1987. 33 1 Left in 2000 and you went to work for a company that was 2 later acquired by the Escher group is that right? 3 A. Yes. 4 Q. You worked with the Escher group for about 21 years 5 rising to chief technical officer, CTO, until your 6 retirement last year, 2021? 7 A. In the last couple of years I actually moved into 8 a consultancy role and did actually work part-time 9 during that -- so I think my job title technically was 10 principal consultant in the last couple of years. 11 Q. Thank you. When you joined the Post Office in 1987 12 I think your employer was the Post Office, a statutory 13 corporation; is that right? 14 A. Yes. 15 Q. And you worked in the IT department of the Post Office; 16 is that right? 17 A. Yes. 18 Q. Did you have normal line management reporting structures 19 within the IT department? 20 A. Yes. 21 Q. You tell us in your statement, no need to turn it up, it 22 is in paragraph 1, that in 1994, so that's about 7 years 23 after you joined, you were "effectively seconded to Post 24 Office Counters Limited"; is that right? 25 A. Yes. 34 1 Q. What did that secondment mean in terms of your 2 employment status? 3 A. To my formal status, none whatsoever. I was basically 4 working on a POCL project but I was still, from a pay 5 and rations point of view, employed by Post Office IT. 6 Q. Was your line manager in the IT department or was it -- 7 he or she within Post Office Counters Limited? 8 A. I had a line manager within Post Office IT but I had 9 very little contact with them to be honest. I was 10 effectively a free operating member of the POCL team. 11 Q. But why were people brought from -- it may seem 12 an obvious question -- the Post Office's IT department 13 into the POCL Horizon programme? 14 A. The way Post Office was structured at that point is the 15 individual businesses didn't have IT functions, there 16 was a central IT function. There were many projects 17 done for the various business units by Post Office IT. 18 Because this was a project obviously with a major IT 19 part of it, they did want to bring in people from Post 20 Office IT. 21 Q. Without that bringing in, did POCL have the relevant IT 22 experience? 23 A. No I do not think they would have done. 24 Q. What, if any, technical expertise was there at POCL 25 management level? 35 1 A. There was a Post Office Counters -- I think it was 2 called an IS strategy information, so information 3 systems rather than IT, looking at the overall strategy. 4 I can't honestly remember what other functions there 5 were. 6 There may have been functions relating to individual 7 projects that were up and running as far as day to day 8 management of those functions within POCL but work on 9 new projects would always have been put out to Post 10 Office IT. 11 Q. Would you know whether at this time, 1994 to 2000, there 12 was any technical expertise at the POCL board level? 13 A. I would not know. 14 Q. Can I turn to the stages of the Horizon programme in 15 which you were involved. So you joined, I think, after 16 the procurement process had been launched in 17 August 1994, the initial very large list of service 18 providers had been cut down to five? 19 A. Yes. 20 Q. At that stage, the five were BT, CardLink, EDS, IBM and 21 Pathway? 22 A. Yes. 23 Q. At the time you joined, the statement of service 24 requirements had already been issued to the five 25 suppliers that I have just mentioned? 36 1 A. Yes, I think so. 2 Q. That was back on 13 April 1995 so pre your joining -- 3 no, that's post your joining, isn't it? 4 A. I can't remember exactly what date I joined the 5 programme. 6 Q. Okay. Can I break down your role over the succeeding 7 six years or so into five parts. These are my 8 descriptions of it. Please agree or disagree with them 9 but it is just to give some structure to your evidence. 10 It is not how matters are arranged in your witness 11 statement. Would it be right that your first role was 12 the evaluation of the five responses, of the five 13 bidders that I have just mentioned, and that resulting 14 in the list being cut to three with BT and the EDS being 15 eliminated? 16 A. Correct. 17 Q. Your second role was within the so-called "demonstrator 18 stage"; is that right? 19 A. Yes. 20 Q. Just explain to us, because it is an odd word 21 "demonstrator stage", what the demonstrator stage was? 22 A. It was a slightly odd phrase for us as well I think. 23 The intention of the demonstrator was to give the Post 24 Office team a chance to look at the three potential 25 service providers, to understand what they are proposing 37 1 in more detail than was within their documentation, so 2 to be able to help them through the process of refining 3 their offerings to us and also to give them the ability 4 to come back to us and ask further questions and 5 clarification. 6 Q. Thank you. 7 A. And at demonstrator -- I think was, they were trying to 8 demonstrate their capability and their solution to us, 9 I think that's where the name came from. 10 Q. The demonstrator phase, is this right, was broken down 11 into streams or strands? 12 A. Yes. 13 Q. And you were the team leader for one of those strands 14 called POCL infrastructure? 15 A. Yes. 16 Q. We will come back in more detail to what that involved 17 in a moment. Your third role was within the evaluation 18 stage and would this be right, in broad terms, it 19 involved scoring the bids against some predefined 20 criteria, and you again had responsibility for part of 21 that process? 22 A. Yes. 23 Q. The fourth role was in the development and assurance 24 stage of the system and that's after the contract had 25 been awarded to ICL Pathway in April and May 1996 and 38 1 then your fifth role was within the live trials that 2 happened thereafter, or various iterations, as part of 3 the acceptance process? 4 A. Yes. 5 Q. Can I turn to the first role -- and I'm going to deal 6 with this very shortly -- the evaluation of the five 7 selected bidders and the narrowing down of that to 8 three. You deal with this very briefly in your witness 9 statement; no criticism there. Is it right that nothing 10 critical emerged at that stage of the process, so far as 11 concerns ICL Pathway and so far as concerns this 12 Inquiry? 13 A. Correct. That was an initial, if you like, sift to 14 reduce down to three and then we were going to be taking 15 those three forward. So I believe the intention was, at 16 that stage, that everybody who went through in those 17 three should be credible, that passed initial hurdles. 18 So they passed those hurdles, I believe, to get to that 19 stage. 20 Q. Thank you. Can I turn straight then to the demonstrator 21 phase or stage and your second role. As we have said, 22 you were the team leader in the demonstrator stage with 23 responsibility for the POCL infrastructure strand. Can 24 we look, please, at WITN05970101. That will come up on 25 the screen, Mr Folkes. 39 1 It is a single-page document, which you exhibit to 2 your witness statement, and it sets out, as the heading 3 suggests, the "Objectives of the Demonstrator Stream", 4 as it is called here. First: 5 "to clarify the requirements with the three 6 suppliers and ensure that they have a valid 7 understanding of these requirements." 8 Second: 9 "to identify deficiencies in the requirements and to 10 feed these back to the BA/POCL requirements team for 11 resolution." 12 Under "Solutions": 13 "to explore and understand the supplier's solutions 14 and to identify issues and risks associated with their 15 solutions (input to Service Provider Risk Register). 16 "to provide a process to allow suppliers to reduce 17 their Risks." 18 Then "Assessment": 19 "to assess the supplier's solutions as input to the 20 overall Evaluation process. 21 "to provide confidence to the stakeholders in the 22 supplier's solutions." 23 Do you agree that this accurately sets out the 24 objectives of the demonstrator stage? 25 A. Yes, I do. 40 1 Q. So, is this right, it was intended to be mutually 2 beneficial, in particular, it was also intended to be 3 beneficial to the service supplier who may ultimately go 4 on and win the bid? 5 A. Very much so and, from the point of view of being able 6 then to get more information but also to be able to 7 raise concerns with us if the requirements didn't make 8 sense or were unmeetable in some way. 9 Q. So it is not just beneficial to you as the clients, it 10 was intended to improve the quality of the service and 11 the system being proposed by the supplier, in the event 12 that they won the bid? 13 A. Yes. 14 Q. This all took place, is that right, in the second half 15 of 1995? 16 A. Yes, primarily, I think, October, November, December 17 1995. 18 Q. Thank you. Can we see what you learned about ICL 19 Pathway in this part of the procurement exercise and 20 look at WITN05970102. 21 This is a summary of activities as part of the 22 demonstrator stage and is a report, we can see, written 23 by you; is that right? 24 A. It is, yes. 25 Q. Who would the report have been sent to? Who was it sent 41 1 to? 2 A. I believe it was input into the overall procurement 3 team, there was a specific team running the procurement 4 and it would have gone, I presume, from there into the 5 evaluation board at various levels within that team. 6 Unfortunately, I can't tell you exactly which 7 individuals received it but the purpose of this was to 8 basically summarise what we had been doing over the 9 three to four months and to feed back the key issues 10 from it. 11 Q. Thank you. So it would have been seen both within Post 12 Office Counters Limited and within ICL Pathway? 13 A. It wouldn't have gone to ICL Pathway because this 14 document contains information on all three bidders. It 15 would have been seen within the PDA, the Programme 16 Delivery Authority, or its -- the team, as then existed, 17 to do the evaluation, to do the procurement. 18 Q. So both the Post Office and the Benefits Agency sides of 19 the house. 20 A. They should have done, yes. 21 Q. Thank you. Can we go over the page, please, and just 22 look at the introduction and the second paragraph reads: 23 "The paper first outlines the structure and 24 organisation of the strand and the general philosophy 25 followed during the life of the strand. This is 42 1 followed by an appendix per service provider, in which 2 their general style is described, together with a list 3 of meetings attended." 4 Then "Background", under paragraph 2 explains when 5 the demonstrator strand was conceived, October 1995. 6 Ran as one of six strands until closure in January 1996. 7 Skipping a paragraph: the original plan did not 8 contain an infrastructure strand, per se, "it would have 9 been partly covered by design assurance", partly by 10 other strands. 11 You say in the last line there: 12 "... the Infrastructure Strand enjoyed the dubious 13 advantage of starting in October from a clean sheet." 14 What did you mean by that? 15 A. I think when the plan was first put together there was 16 no concept that we needed to look at the infrastructure. 17 The programme was based around the idea of a number of 18 different services and some of those services were 19 applications, such as BES or EPOSS or APS, within the 20 Post Office side, but obviously all those needed to run 21 on a platform and what we were trying to procure was 22 an overall platform for POL to -- Post Office to go 23 forward with and the discussions that we had in the 24 programme was that you needed to look at that platform 25 as a whole. 43 1 Q. Got it. 2 A. It was no good looking at -- we weren't going and buying 3 an Automatic Payment Service and separately a BES 4 service and separately an EPOSS service. They were all 5 going to be running on the same hardware provided by the 6 service provider, the same underlying platform services, 7 the same comms network, the same central services. So 8 it was to look at that -- the techie side, if you like, 9 rather than the application side. 10 Q. I understand. Can we go forward to page 5, please, and 11 look under "Approach" in paragraph 5, and 5.1 12 "Demonstrator Meetings", to see what happened. You say: 13 "The demonstrator process was based around a series 14 of full day meetings with each service provider, with 15 each strand being allocated a specific day of each week 16 for each supplier ... as part of the planning at the 17 Introductory meetings. 18 "[They] were generally held at the service providers 19 main offices ... some were located at specific 20 subcontractors sites if this facilitated specific 21 demonstrations ... In addition, a two-week period in 22 December 1995 was reserved for 'site visits' ... this 23 being used for a combination of reference site visits 24 and presentations/discussions with overseas 25 subcontractors." 44 1 You set out the format of each meeting. So is that 2 the approach taken? 3 A. It was, yes. 4 Q. Can we go forward to page 15 to see what you say about 5 ICL Pathway. I should say that, in between page 5 and 6 page 15, there is information about the other bidders. 7 I'm concentrating on Pathway. 8 This is divided into two parts, what you say about 9 it. If we just look at the headings first, C.1 is 10 "Technical Infrastructure Area", and then, over the 11 page, please, at the foot of the page, "Support Services 12 Area", yes? 13 A. Yes. 14 Q. If we go back to the first page, please. Thank you. 15 "Technical Infrastructure Area", what does "Technical 16 Infrastructure Area" mean? 17 A. So this basically meant everything around the technical 18 platform that Pathway would be providing for the 19 application services to run on. So this would include 20 everything from the hardware in the office, the 21 communications networks to the centre, what systems 22 would be provided at the centre and all, if you like, 23 the common elements that were not part of the individual 24 applications. 25 Q. I understand. You say this: 45 1 "Meetings with Pathway took place in the meeting 2 rooms at their offices Feltham. Unlike the other two 3 suppliers, the meetings were fronted by sales orientated 4 rather than technically orientated staff -- initially 5 Liam Foley, and then Martin Johnston -- and these 6 representatives also took the notes/actions from the 7 meetings. 8 "Numbers at the meetings varied", you set out some 9 core members of the team with others being brought in. 10 In the third paragraph, you say: 11 "The meetings were characterised by less structure 12 and less evidence of preparation than with the other two 13 suppliers. Some presentations were given, however these 14 were fairly informal with very few prepared slides, with 15 diagrams drawn on a flipchart when required. 16 "Papers were initially hard to extract from Pathway, 17 and although this problem did ease up to a certain 18 extent during the demonstrator ... a significant amount 19 of chasing was required to career the outstanding 20 documentation at the end of the phase. Papers 21 themselves varied in quality and detail, with less 22 evidence of internal review prior to issue." 23 Then scroll down, please: 24 "Towards the end of the demonstrator phase, Pathway 25 started taking a fairly robust attitude on risks, with 46 1 the appearance of their Risk Director (Martyn Bennett) 2 at the start and finish of each meeting to check on the 3 status of risks and actions. Although Martyn took 4 a fairly aggressive attitude towards the demonstrator 5 team, his prime raison d'être seemed to be to ensure 6 that the Pathway staff produced suitable responses. 7 Despite this additional focus, adequate risk responses 8 were still difficult to obtain and a number of risks 9 required repeated iterations of responses before 10 clearance could be recommended. 11 "In the cross stream meetings, Pathway again took 12 a more sales-orientated approach, with less solution 13 content than with the other two suppliers. This was 14 being evidenced by the demonstration of the somewhat 15 irrelevant Household Budgeting Scheme, and of 16 demonstrating putting demo team photos on cards, rather 17 than showing a prototype of their solution to genuine 18 requirements." 19 You are identifying maybe five or six problems with 20 the way in which they were demonstrating to you? 21 A. Yes. What we wanted these meetings to be was a genuine 22 substantive interaction back and forth between the 23 service provider and our team. We felt that that worked 24 with the other two service providers, as far as we could 25 expect. I think it was six meetings over a number of 47 1 weeks. 2 What we found here with Pathway is it was the team 3 very much as a sort of sales event. They were run by 4 a sales team rather than a technical team and, you know, 5 they hadn't prepared as much in the way of technical 6 documentation for us, or diagrams or whatever, and that 7 last comment in there was it felt very "salesy". They 8 were showing that you could give customer's cards with 9 photos on so they brought in a card printer and took our 10 photos and put the photo on a plastic card. 11 That might work from a sales point of view. From my 12 point of view, that was no good whatsoever. I know you 13 can put photos on cards. I wanted to understand their 14 technical solution. 15 So they seemed to be operating at a slightly 16 different level and they almost saw these as a sales 17 event that kind of had to be done, rather than 18 opportunity to actually work with us. And going back to 19 what you said earlier, the intent here was that it was 20 going to be of mutual benefit to help us understand them 21 but also to allow them to improve their solutions. 22 Q. To what extent does what we read here about Pathway's 23 approach reflect the remaining time that you spent 24 working with Pathway over the succeeding few years? 25 A. Again, it would be simple to say it does reflect it. 48 1 I think it probably reflects it for slightly different 2 reasons, in that when we moved forward into the 3 contract, it was the PFI nature that I think caused 4 a total difference in opinion. In this case, I'm not 5 quite sure why they took this approach. From our point 6 of view, it was equally as annoying because we wanted 7 to -- we genuinely wanted to help all three service 8 providers to come up with the best offerings to the Post 9 Office and BA so that we could then pick the best one 10 from them. 11 Q. So it would be dangerous to say that what we read here 12 demonstrated behaviours or an attitude of mind by the 13 company that we then see played out over the succeeding 14 years in its dealings with you? 15 A. In hindsight, I can see similarities but, as I say, 16 I think they are probably for different reasons. 17 Q. In the paragraph starting "Towards the end of", you say: 18 "... Pathway started taking a fairly robust attitude 19 on risks ..." 20 What does that mean, please? 21 A. Okay, so one of the outputs from the demonstrator phase, 22 or one the tools we had, is we could raise risks against 23 each service provider. 24 Just to be clear, risk was, basically, we're raising 25 a potential issue and giving them a chance then to 49 1 address it. So the fact that a risk was raised, if it 2 was then addressed it wouldn't count against them, but 3 it gave us a formal mechanism to say "Look, you talked 4 about this, either we don't believe it or we don't think 5 it is going to work or it doesn't meet a requirement", 6 whatever, "we are going to raise a risk against it", 7 that put it down on paper as a risk. 8 Those were then shared with them and they were then 9 given the opportunity to address those risks. And the 10 idea was they could then come back with a risk response 11 which we could then evaluate and that might say "Yes, 12 that risk is fine, it has been cleared", or it would be 13 a risk, if it wasn't cleared, that would then go forward 14 into the evaluation and then actually taken forward with 15 the Pathway risks into the real projects. 16 So what Martyn Bennett was -- I think they sussed at 17 this point that if they didn't take us seriously on 18 risks they were going to count against them because it 19 wasn't just the individual's strands who were being -- 20 that they were meeting with. If I raise a risk, that 21 risk went into the service provider -- 22 That risk went into the SPRR, the service provider 23 risk register and would then have full visibility and 24 count against them. So I guess, at this point of view, 25 Martyn was doing his job, which was to minimise the 50 1 number of risks. 2 At some points, it appeared to us that he was not 3 keen on us not raising risks but he was also, in that 4 line here, I think he was raising that and that seemed 5 to be to ensure that Pathway's staff produced suitable 6 responses; he wanted them to go and clear the responses 7 because he was going to be marked down if he ended up 8 with 20 risks at the end of it. 9 Q. I understand. Can we go over the page please and look 10 at the bottom half of the page, "Support services area". 11 Can you explain what support services area is please? 12 A. Okay. So when we set up the POCL infrastructure stream, 13 there was a second -- there was a piece of work that 14 didn't really have a home and that was the support 15 services such as helpdesks, such as field engineering, 16 I think such as installation, that didn't really have 17 a home until we invented POCL infrastructure and then it 18 was added to my remit within POCL infrastructure. 19 Because support services, Helpdesk, et cetera, 20 wasn't my speciality, I actually got in an experienced 21 person from POCL who sort of led on this -- a guy called 22 Steve Grayston, who led on the support services side. 23 So that was very much looking at helpdesks -- 24 Q. I think I have the description, thank you. You say that 25 Stephen Muchow was the Pathway consortium representative 51 1 who dealt with the POCL support service aspects -- 2 I think we are going to hear from him in a later 3 phase -- and you say "The impression was given that the 4 methodology", who was that impression given by? 5 A. Given by Stephen Muchow, I think. 6 Q. "The impression was given that the methodology by which 7 the procurement was being conducted was unnecessary and 8 time consuming/costly. Meetings were conducted in 9 a more ad hoc manner. Papers and reference documents 10 were not easy to obtain and, when received, were less 11 comprehensive than anticipated. Discussions although 12 fruitful were less flowing than expected. 13 "A site visit was made [at the ICL facility in 14 Havant]. This was well conducted with appropriate 15 consortium personnel available for 16 questions/information." 17 Now, in the course of this stage of the process you, 18 as you have said, held a series of meetings with ICL 19 Pathway, especially in November 1995 and I think notes 20 were taken of the meeting which were turned into meeting 21 reports; is that right? 22 A. Yes. 23 Q. I just want to establish the date and the structure of 24 them first please. Can we look please at WITN05970106. 25 This is a meeting report of 1 November 1995. We can see 52 1 the attendees from the Benefits Agency and POCL, it was 2 you and Mr Grayston, who you just mentioned, and then 3 from the supplier side -- from ICL's side on the 4 right-hand side. 5 A. Yes. 6 Q. Did you write these? 7 A. I wrote these yes. 8 Q. If we just go through them to put them in evidence. Can 9 we go to WITN05970107 please. 10 This is a report for the 8 November we have just 11 looked at 1 November, this is 8 November. And then same 12 reference but 0108. 13 Same reference -- sorry, that's on 15 November 1995. 14 Same reference but 0141, that is for 22 November. Same 15 reference 0109, 29 November 1995. Same reference 0142, 16 that is for a meeting on the 11 and 12 December 1995, 17 I think in Boston; is that right? 18 A. Yes. 19 Q. Then, lastly, same reference 0139, 11 January 1996. So 20 those are contemporaneous records that you took, turned 21 into meeting reports of your contact with, at these 22 meetings, ICL staff? 23 A. Yes. Just to say, we were having meetings with all 24 three service providers on Tuesday, Wednesday, Thursday 25 and generating copious notes and, I can't remember 53 1 whether it was that evening or the Friday, we spend the 2 time turning these notes into permanent records and then 3 taking forward the individual items within there, 4 including, in many cases, either raising risks or 5 closing down risks. 6 Q. That's why we see, for example, the dates 1 November, 7 then seven days later on 8 November, then seven days 8 later on 15 November, et cetera. 9 Can I look at some of the thing that were said in 10 the course of three of those meetings reports. Can we 11 go back to same reference, 0106, please. 12 This is the meeting note of 1 November 1995. Can we 13 look please at the second page at paragraph 13. Thank 14 you: 15 "Counter hardware. Pathway believe -- 'no problem 16 freezing tech base during the rollout' (if 17 12 month-ish). Lengthy discussion, [especially] around 18 POCL wanting to test [equipment]. Pathway expecting 19 very hands-off approach -- and practicalities of 20 expecting that stability. All down to Pathway's 21 agreement with sub-cons. (!)" 22 Can you tell us what that means please, that 23 paragraph. Decode it for us? 24 A. Okay. The rollout was going to be -- would have taken 25 12 to 18 months to rollout to 20,000 offices and 40,000 54 1 or so terminals. One of the questions that we asked, 2 probed, was how they were going to actually manage doing 3 that rollout, as in, do they buy 40,000 bits of kit and 4 put them in a warehouse and roll them out or are they 5 going to be taking them on month by month? Computers 6 changed, technology changes, in particular, firmware 7 versions in PCs change. 8 It is quite possible that at the start of 18 months 9 the end of 18 months, the hardware may have changed and 10 that would affect things like testing and field support 11 etc. We weren't trying to manage it, we wanted to 12 understand how they would manage it. 13 Q. What was the sentence "Pathway expecting very hands off 14 approach" intended to convey? 15 A. They were expecting BA/POCL to be hands off. 16 Q. What do you mean by hands off, what did they mean by 17 hands off? 18 A. So they were not expecting us to have any particular 19 interest in this or any involvement in this. So their 20 view would be -- I think it is their job to manage the 21 hardware, what are you worrying about? 22 Q. Thank you. Can we turn to same reference 0107, please. 23 These are the meeting notes -- or the report of the 24 meeting for 8 November 1995. Can we turn to page 5, 25 please. Can we look at the fourth and fifth bullet 55 1 point entitled "lack of cohesion" and you write: 2 "Lack of cohesion between the people at the meeting, 3 must be doubt over ability to manage project if this 4 interface to their customer is so weak. 5 "General problem that there is no documentation 6 about the system, and late arrival of Mike Murphy and 7 his contradiction/clarification re earlier explanations 8 call into severe doubt the knowledge of the consortia 9 about what they are proposing and then how they may 10 develop it/support it in any timescale." 11 Can you explain what you meant by those two bullet 12 points, please? 13 A. Yes. So this meeting, we were looking at the Riposte 14 software that Pathway were proposing. Riposte software 15 which was already in use in An Post the Irish Post 16 Office. Mike Murphy was the CEO and I think owner, part 17 owner, of Escher, the provider of Riposte at the time 18 and Mike came to this meeting. 19 The observation generally through this note is, 20 there seemed to be a lack of join up between the Pathway 21 people and what Mike Murphy was saying. And, you know, 22 the concern here was that we were there as the customer, 23 we were expecting Pathway to provide a single voice. We 24 were more than happy to have an expert/experts from 25 Escher there, that what's what we wanted but what we 56 1 found was a lack of join up as to what it was that 2 Pathway were providing. The whole purpose of the 3 demonstrator phrase was to try and get clarity as to 4 what they were providing. 5 Q. You make a bigger point, "must be doubt over ability to 6 manage project if this interface to their customer is so 7 weak". You are pointing towards the entirety of the 8 project there, is that right? 9 A. I think what I meant in that point was, given the lack 10 of cohesion between this team, in what was probably one 11 of the more important meetings they would have had -- 12 you know, when I wrote these notes I was thinking "how 13 would this work in reality on the whole projects". Now, 14 I went on to say that there seemed to be contradictions 15 and the comment there that there seemed to be doubt 16 about the knowledge of the consortia, ie Pathway 17 overall, and what it was that they were proposing. 18 Q. Thank you. Then lastly in this selection of minutes, 19 can we go to the same reference 0109 please. Where this 20 is the meeting report of 29 November 1995. Can we go 21 forward to page 5 please. Under the heading "Riposte" 22 in paragraph 5, look at the second bullet point, you 23 say: 24 "Went over the failure in the walk through case in 25 tedious detail (again) with Pathway tripping over 57 1 themselves, and Martin Johnson trying (badly) to show 2 that he understood something. Pathway seemed keener to 3 rubbish the specific example we had come up with, rather 4 than addressing the issue. Seemed to miss the point 5 that Mike Murphy had acknowledged it and did have 6 an answer (using strong identity -- see earlier) but 7 Pathway were unable to explain." 8 Can you decode that for us? Explain what you were 9 referring to there? 10 A. I can't remember what the specific case was but we had 11 come up with, "well, how will this work". The Riposte 12 system, if I can just digress for a moment, involves 13 software running on each one of the terminals in the 14 office and these terminals, they can replicate data 15 between each other and replicate data up to the data 16 centre and that's the whole benefit of this and the 17 whole way the system operates. We, I think, had come up 18 with questions about failure cases, what would happen if 19 that became disconnected, or that became disconnected? 20 And what we wanted to tease out is (1) how did these 21 failure cases happen, but also, what became clear is 22 that the Pathway board didn't necessarily understand it. 23 In this case Martin Johnson -- who I think was a sales 24 person, who was sort of, no disrespect meant to him, but 25 he was in these meetings to sort of manage the meeting 58 1 rather than provide technical information -- went off 2 trying to explain how Riposte works; which wasn't 3 a great success from what I see here. When we started 4 raising issues on failure conditions they seemed to be 5 keener on rubbish the specific examples we put forward 6 rather than trying to address it. 7 Q. Can we go over the page, please -- sorry. 8 A. In a population of 40,000 terminals, if it can go wrong, 9 it will. You know, it is not like -- I am sure the IT 10 guys with their systems here will probably tell you the 11 same with 100 terminals. In a big Post Office estate, 12 failure of hardware and network counters(?) it was 13 always going to happen, therefore you needed to ensure, 14 especially with a distributed system, that there was 15 adequate coverage of the failure cases and that the 16 system would work. What we had here was that the 17 examples appeared not to be taken seriously rather than 18 being addressed. 19 Q. In that sentence you said "especially with a distributed 20 system". What do you mean by a distributed system? 21 A. Okay. So there are -- in the Riposte system there's 22 persistent storage on every terminal and the great 23 advantage of this, it gives the ability for the terminal 24 to operate even if the communications to the centre is 25 offline. But a number of the other solutions, including 59 1 I believe Horizon online, which Post Office then moved 2 to and also from, I believe, CardLink and IBM, they were 3 reliant upon online solutions. That would have meant 4 that if the communications to the centre went down for 5 any reason, that there would be cases where the counter 6 clerk just wouldn't be able to serve and the ability to 7 serve, even in cases of failure such as a network going 8 down, was important to the Post Office. And that was 9 one of the perceived advantages of the Pathway solution, 10 that it was distributed and therefore as long as the 11 terminal was up and running, they could still serve. 12 Q. Thank you. Over the page please. Can we look at the 13 second bullet point from the bottom please. You say: 14 "Another extremely frustrating, contradictory day. 15 Pathway seem afraid to admit that they are changing 16 Riposte. On [the] one hand it is of concern as the 17 changes reduce the value of the reference sites/track 18 record, on the other it is good at showing that they 19 recognise that the product is not perfect and may need 20 changing for the higher volume environment that they are 21 proposing ... However, the secretive and ill-informed 22 attitude is damaging credibility and amplifying our own 23 doubts over the viability of the product." 24 We have looked at three examples of the things that 25 you said in meeting reports on 1 November, 8 November 60 1 and 29 November 1995. 2 Was the kind of thing that you were saying here, fed 3 back into those who had power to make a decision over 4 the award of this contract and if so how? 5 A. It was fed up in the summary document, particularly it 6 was fed up in the risks. So we had specific risks. 7 The risks had very high visibility. Risks on, for 8 instance, the scalability Riposte, the relationship with 9 Escher, the ability of Pathway to manage the 10 relationship with Escher and that those risks -- I think 11 I go through a list of them in the witness statement -- 12 those risks were visible all the way up to the 13 evaluation board. 14 Q. I was about to say, to which people within POCL were 15 these meeting reports sent or distributed? 16 A. These meeting reports that I had here, I think would 17 have only been distributed within the demonstrator team. 18 These were obviously very detailed contemporaneous 19 material. 20 The risks that came out of these went up into the 21 procurement team, to the service provider risk register 22 and they would have been visible at the level of the 23 evaluation team management and the -- I can't remember 24 the names on all of those but there is a document which 25 records the outcome of the evaluation that has a list of 61 1 the personnel who were in the -- on the evaluation team. 2 Q. Notwithstanding what was written here, we know the five 3 bidders were narrowed to three. Did you express a view 4 at that time about Pathway being amongst that group? 5 A. Sorry, these meetings here were when we were already at 6 three. 7 Q. Thank you. Can we turn, then, to your third role which 8 was, I think, the evaluation stage and you say that 9 there was a part it codenamed Amazon. 10 A. Yes. 11 Q. Amazon was, more specifically, an evaluation of the 12 three bidders responses -- paper responses to the rather 13 extensive invitation to tender document; is that right? 14 A. Yes. 15 Q. Would this be in February and March 1996? 16 A. I believe so, yes. 17 Q. Did it involve the team, of which you were a part, 18 working away from the office, in particular, for 19 an intensive two week session? 20 A. Yes. 21 Q. Looking at it in high level summary, were there two 22 elements or parts of it: firstly, a financial and risk 23 transfer evaluation and, secondly, a business and 24 technical evaluation? 25 A. Correct. I think it was the second of those that I was 62 1 involved in. 2 Q. Now, I think the evaluation took place by reference to 3 a combination of monetary and nonmonetary factors, is 4 that right? 5 A. Yes, although the part that we were involved in was 6 purely on the qualitative rather than the quantitative. 7 Q. Can we look at those please at POL00031154. 8 You can see this is a document of which you are not 9 the author dated 11 March 1996, "Supplier scores in 10 respect of value factors". You are familiar with this 11 document I think? 12 A. Yes. 13 Q. Can you explain, in general terms, what it is? 14 A. Okay. So the procurement team -- and Derek Selwood was 15 a member of that team -- had come up with a scoring 16 model. 17 The model had a fairly complicated mapping of, 18 "value factors" they were called, in a number of 19 different categories. 20 I believe there is a table in there that explains 21 what those are. I can't remember them individually. 22 Q. Can we go to page 7 of the document please. 23 If we just look at the whole thing. Is that the 24 table of the value factors that you were just speaking 25 about? 63 1 A. Yes. 2 Q. Then if we go to page 9 please. We can see the -- if we 3 can display page 10 at the same time please. We can see 4 the two teams, the "contracts stream" on the bottom of 5 the left-hand page and the "demonstrator stream" on the 6 right-hand page? 7 A. Yes. 8 Q. Of which you were a member and the group leader for, 9 again, POCL infrastructure; is that right? 10 A. Yes. 11 Q. Can you just tell us what, before those two streams, the 12 programme review panel was, the membership of which is 13 set out in the top left-hand side of the page? 14 A. I think the programme review panel was, if you like, the 15 senior team providing the evaluation. They had 16 representatives from each one of the streams. So you 17 have got on there one or two representatives from the 18 demonstrator team, one or two representatives from the 19 contracts team. You also had people on there who were 20 not part of the demonstrator, for instance Bob King, who 21 was a Post Office Counters manager but he wasn't part of 22 the programme as such. 23 Q. Thank you. 24 A. The demonstrator team was doing detailed work that then 25 fed into that and I presume the contracts stream 64 1 likewise. 2 Q. Can we go forwards to page 11, please. This is annex C 3 of the document. If we can look at the whole page to 4 start with, please. I think this breaks down the ten 5 value factors that we looked at earlier. In fact, we 6 only displayed 1 to 7 of the value factors. 7 Then, by reference to a series of topics identified 8 in the left-hand column, it shows us by a tick the 9 likelihood of those value factors having relevance to 10 the issues identified; is that right? 11 A. Yes, I don't think I would use the word "likelihood". 12 They were the ones that -- we were scoring, I believe, 13 based upon what's in the left-hand column by the 14 different functional areas of the solution against 15 criteria set by the value factors and some of those -- 16 some of the value factors weren't deemed relevant to the 17 area, hence the lack of a tick. 18 Q. For your part of this stream of work, looking at the 19 list down the left-hand side, which of the identified 20 headings and then subheadings were of relevance to you? 21 A. I will be honest, I can't remember whether we just did 22 our own area or whether, as a team -- I have a feeling 23 as a team, the whole team went through it but maybe we 24 led on an area. Certainly, the POCL infrastructure 25 section on there would have been the area that I would, 65 1 if you like, led upon. 2 Q. That included "OP". What does "OP" stand for? 3 A. Okay. So the vision for the programme had a number of 4 services and one of the services was the office platform 5 service, OPS, office platform being, at its simplest, 6 the PC in the office and the devices attached to the PC 7 in the office and, in that case, the middleware which in 8 this case would have been Riposte, the firm of 9 suppliers, would have been whatever middleware they were 10 proposing. 11 Q. Can you just explain to the Chair what middleware is, 12 please? 13 A. Okay. Typically a software system has a stack of 14 different elements to it. At the top of that stack you 15 would have applications, bottom of the stack you have 16 the hardware. Middleware, as it is, suggests is what 17 goes in the middle. It tends to be software that 18 provides a technical function, such as a database or, in 19 the case of the Pathway solution, message replication 20 system. It doesn't necessarily have any sort of user 21 interface. The counter clerk user doesn't know they are 22 interacting with it but it is sitting there providing 23 a vital function. 24 Q. So the type of software that provides services to 25 software applications, other than the operating system? 66 1 A. Correct. Far better way of putting it, thank you. 2 Q. "WAN", please? 3 A. Wide area network. 4 Q. What's the wide area network? 5 A. So the wide area network is what was going to link the 6 individual post offices to whatever central system which 7 the service provider had. In the Pathway case, that was 8 using ISDN in the majority of the cases. 9 Q. TMS, transaction management service? 10 A. That was the central systems that the service provider 11 would provide, which, in Pathway's case, would have 12 included the central correspondence servers and other 13 sort of mainframe type systems behind them. 14 Q. In order to score these issues listed under POCL 15 infrastructure, that would require the demonstrator 16 team, including you, to be able to judge satisfaction or 17 compliance with the value factors at a detailed 18 technical and granular level; is that right? 19 A. Yes. That was done by reference to the responses that 20 the service providers gave to the information to tender. 21 So they provided the -- each one provided their bids and 22 they had to respond, I believe, to each individual 23 requirement and then we had to score those individual 24 requirements. 25 Q. Thank you. Can we go to paragraph 34 of your witness 67 1 statement, WITN05970100, at page 11, paragraph 34. You 2 tell us: 3 "One key aspect of PFI [private finance initiative] 4 was that requirements had to be presented as high level 5 'Output-Based Specifications' so as not to constrain the 6 Service Providers, that is stating the 'what' but not 7 the 'how', and at the level of the Service rather [than] 8 a system. This was contrary to most experience where 9 specifications would be far more detailed and include 10 some 'how'." 11 You are making the point here, is this right, among 12 a series of other points about the effect at this stage 13 of the contract being a PFI? 14 A. Yes. 15 Q. How could the exercise that we have just looked at be 16 undertaken or how could it be completed without detailed 17 insight into the "how"? 18 A. The requirements had to be presented as high level 19 output-based specifications. We weren't allowed to tell 20 them how to do something. The service providers had to 21 respond in their tender documentation and tell us to 22 a level how they were going to meet that requirement, 23 and the purpose of the demonstrator was also that we 24 gained understanding of how it was going to be done. So 25 what we were scoring was their responses to the 68 1 invitation to tender. 2 Q. Without knowing what sat underneath the responses? 3 A. They had to provide sufficient information in the 4 response for us to be able to score it. That wasn't 5 going to be down at the level of detailed design because 6 at this stage they had not done detailed designs but 7 they had to provide adequate information and, if they 8 didn't provide sufficient evidence, then they would have 9 got a low score. 10 Q. You tell us in paragraph 35, if we can continue: 11 "The other key aspect of PFI was that of Risk 12 Transfer. Whilst there was an explicit and well-defined 13 attempt at Risk Transfer for certain types of Benefit 14 Payment Fraud by BA, there were of course many Risks 15 which could not be transferred to a Service Provider, as 16 POCL has found to its cost over the years. 17 "36. Outsourcing a major part of your core business 18 (and the transaction processing/counter system for 19 a Post Office is about as core as it gets) does not 20 remove the risk to that business if the counter system 21 (or service) does not perform. Paraphrasing somewhat, 22 the ethos was that risk was transferred to the Service 23 Provider, and so we should not worry about it. If they 24 failed it was at their cost. I think this was 25 a fundamental issue -- whatever idea there was of Risk 69 1 Transfer in specific areas, POCL still needed assurance 2 in the quality and fitness of the service being 3 developed and provided by the Service Provider, to 4 manage the risk to their business." 5 What you have said there in paragraph 35/36 is said 6 by you in the context of the transfer of fraud risk to 7 ICL Pathway but does the point apply more broadly, 8 ie that ICL Pathway were saying that POCL should not 9 worry about the project because, if it failed, then it 10 was at their cost? 11 A. Simple answer to that is yes. Can I just say 36 wasn't 12 specifically about fraud risk. It was -- what I was 13 trying to say there was the PFI ethos was all about risk 14 transfer and what I feel in 36 is that some risks can be 15 transferred, for instance, certain risks regarding 16 benefit encashment were transferred but the risk that 17 the service provider gets it wrong and screws your 18 business can't be transferred. 19 And that was the whole basis on which we felt, in 20 the assurance team -- we have jumped forward a bit 21 here -- in the assurance team, that we did need to, if 22 you like, know what was in the box. Waiting for the 23 service provider to get it wrong and then they then take 24 a financial penalty, it wasn't going to help POCL. What 25 we wanted to do was to understand what Pathway were 70 1 doing and make sure they were doing it in the right way 2 so that we ended up with a successful solution. 3 Q. You are saying that ICL Pathway's ethos was that you -- 4 that's POCL -- should not worry about it and you are 5 saying this in your statement at a point where the 6 contract was still a PFI one involving the Benefits 7 Agency. Did the same approach still apply when the 8 Benefits Agency had pulled out: this was not a PFI 9 contract and was instead a much more standard design and 10 supply contract? 11 A. In 1999, bearing in mind I left at the start of 2000, so 12 I don't know how it went after that -- in 1999, I think 13 it was still very much in that ethos. I think the key 14 point, though, is the damage was done by then. We had 15 gone through from when Pathway started work in 1996 16 through until the point that the contract was 17 renegotiated in the second quarter of 1999, under the 18 PFI cloud. 19 And suddenly saying this is not a PFI contract any 20 more, we didn't suddenly get access to everything we had 21 not had access to over those three years and things that 22 may have happened during those three years didn't 23 mysteriously go away. 24 SIR WYN WILLIAMS: Mr Beer, if I may, I think I need 25 a five-minute break. 71 1 MR BEER: Sir, of course, and in fact I fortuitously was 2 about to suggest one because of the transcribers have 3 suggested that we should go for an hour or so. 4 SIR WYN WILLIAMS: Well, then, if that's the case, are we 5 having a more traditional break or just five minutes? 6 MR BEER: Just five minutes, please, sir. 7 SIR WYN WILLIAMS: Thank you. 8 MR BEER: Thank you very much. 9 (12.07 pm) 10 (A short break) 11 (12.14 pm) 12 MR BEER: Can we turn, Mr Folkes, to paragraph 37 of your 13 witness statement, which should be on the screen. You 14 say: 15 "As the Project progressed, it appeared that the 16 effect of PFI was that we were expected (by the [Service 17 Provider]) to treat the underlying solution as a 'black 18 box'. The Service Provider's job was to ensure it 19 created the right outputs but the contents of the box 20 were not available for scrutiny -- either how it worked 21 or how it was being built. As I will cover later, this 22 had major effects on POCL's ability to gain assurance on 23 the solution and to inform later activities." 24 This idea here that ICL Pathway expected you and 25 Post Office to treat the system as a "black box" is one 72 1 that you return to in the course of your statement. Are 2 you using the term "black box" there as a term of art, 3 not in the sense that a layman like me might understand 4 it, to refer to and only to refer to a flight data 5 recorder on an aeroplane? 6 A. It is not a flight data recorder. A black box, I think, 7 is something that you can't see inside. 8 Q. So it's a term of art, would this be right -- I'm going 9 to have another crack at a definition here -- it is used 10 in science, computing and sometimes engineering to refer 11 to a device, system or object which produces useful 12 information without revealing any information about its 13 internal workings? 14 A. Yes. 15 Q. Is that the sense in which you're referring to "black 16 box"? 17 A. What I was meaning here is the approach that the service 18 provider took was that the solution was a box that we 19 couldn't see inside. It was there to take those inputs, 20 provide various outputs. They had a set of 21 requirements. Their job was to make sure it met the 22 requirements but we weren't permitted to either see how 23 it worked inside that box or how they had built the box. 24 And I noticed the word used in another document, it is 25 a matter of transparency. The black box is not 73 1 transparent. 2 Q. You say, in the last sentence, that this: 3 "... had major effects on POCL's ability to gain 4 assurance on the solution and inform later activities." 5 Why was that? 6 A. Okay, so we were set up, post award of contract, as 7 a team whose role it was to try and provide design and 8 technical assurance of what Pathway were producing, if 9 you like to provide an oversight of what Pathway were 10 doing to provide a good feeling back to the sponsors, to 11 feed issues back to the sponsors and to make sure 12 Pathway didn't head off in the wrong direction for 13 whatever reason. 14 What we found were that Pathway were very reticent 15 to let us actually get hold of information as to how the 16 system worked internally, in particular denied us access 17 to design documentation. Now, that may have been 18 because such documentation didn't exist. At the time, 19 we assumed it was because they were playing what I call 20 later in the document the PFI card, they didn't want our 21 interference in it and didn't want to share that 22 documentation with it. 23 Because we couldn't see how it was going to be 24 produced and how it was going to work internally, we 25 could that the then go back to the sponsors and say 74 1 "Yes, what Pathway are doing on EPOSS or APS, or 2 whatever, is good and meets our requirements". 3 What I mean by later activities is we weren't trying 4 to trick Pathway, we were trying to make sure it was 5 going in the right direction and then make it easier 6 when we got to acceptance to make sure that what had 7 been produced was going to get through acceptance. 8 Obviously, getting to acceptance and not knowing 9 what was inside the box made it far harder to carry out 10 acceptance properly. 11 Q. Looking back at the matrix that we viewed before the 12 break in annex C, wouldn't it make it very difficult or 13 impossible to undertake any qualitative analysis of the 14 type required by that exercise? 15 A. It would but, bearing in mind what we are talking about 16 here is the comments on PFI is what happened after 17 award. During the evaluation process and during the 18 demonstrator phase, in particular, the service providers 19 had to produce their responses. We had a mechanism 20 through the risk creditors to be able to say "Hey, we 21 are worried about this or that", and they had to produce 22 documents. If they didn't, it counted against them. 23 So they had a positive incentive to respond to us. 24 Once we got to assurance they had a positive incentive 25 to go as far as saying obstruct us in some cases. 75 1 Q. We would come onto what you say about ICL Pathway's 2 obstruction a little later. 3 Can we look at paragraph 38. You say: 4 "Put informally, the approach seemed to me a case of 5 'Trust Me, I'm a Doctor' -- having told them at a high 6 level what we wanted the Service to do, we were meant to 7 trust them as the experts to create and run the Service. 8 We would have Acceptance at the very end, but we would 9 have no visibility of what was 'in the box' or how it 10 had been built, and only be able to perform Acceptance 11 based on those output-based specifications." 12 You are obviously saying here that the "doctor" was 13 ICL Pathway saying, "You need to trust us, we are 14 doctors, we are experts in information technology"? 15 A. Yes. 16 Q. Did you have this unease at the time or is this 17 something that you have thought about on reflection 18 20 years later? 19 A. We had unease during the programme that we were -- our 20 job was to do assurance and we couldn't do the job. And 21 I know we reported this up and it is documented in 22 a number of places, all the way up to the top level 23 decisions, that we had not been able to get assurance on 24 the state of the project. 25 Q. When you say reported up to the top level decision, you 76 1 mean to the POCL board? 2 A. I can't say it was to the POCL board. I haven't had 3 access to everything they did but, certainly, if you 4 look at the documents from middle of 1999 around 5 acceptance, there's statements in there around the lack 6 of assurance that we had been able to get and that's the 7 culmination of what we had been trying to do from 8 1996/7, all the way through. 9 Our intention was not to try and trick people. We 10 all wanted this thing to get through. What we wanted to 11 do was to get assurance as to the direction in which 12 they were going and, in cases where we were allowed 13 access, I think generally we came away with quite a good 14 feeling on a number of them and some of that was around 15 some of the deep, technical infrastructure. It was the 16 areas where we didn't have access that it appears that 17 we may not have had access, essentially because there 18 were problems in those areas. 19 Q. Is that why you explain them playing the PFI card? 20 A. Yes. 21 Q. What do you mean by "playing the PFI card"? 22 A. What I mean is that, in a number of cases where we tried 23 to get access to documentation, the answer would always 24 come back, "No, it is PFI, it is not appropriate for you 25 to have that", as we heard from Terry Austin. It wasn't 77 1 appropriate for the contract, or whatever the words he 2 used. No, the contract -- I think part of the problem 3 here was the contract didn't support us getting this 4 level of visibility. The service provider could have 5 given it to us but chose not to, which is why, in my 6 statement, I do say I think the way the contract was put 7 together was the cause of some of these problems. 8 Q. You have described it as -- 9 Sorry, that can come down, that witness statement, 10 now. 11 You have described it as "playing the PFI card", 12 which has a pejorative element to it, as opposed to 13 "doing that which is required under the PFI contract". 14 Why do you say that they were "playing the PFI card?" 15 A. I say that because it was blocking the task that we were 16 trying to do and that we, by the nature of our roles, 17 were tasked to do. I think Pathway would have made 18 documentation available to us, could have involved us 19 more, could have been more open and transparent with it 20 but they, in a number of areas, you know, forcefully did 21 block access. It didn't just seem to be a passive 22 thing, it seemed to be actively preventing access, which 23 is why I used the language I did there. 24 Q. Can we go back to the evaluation stage and look at 25 POL00028152, please. 78 1 Looking at the front page, this is the evaluation 2 team's final report, dated 28 April 1996. Did you see 3 this at the time? 4 A. I can't remember if I did or not. There certainly are 5 parts that our work went into but I can't remember if 6 I actually had a copy of this at the time. 7 Q. Can we turn to page 15, please, and look at the bottom 8 of the page, please, under "The Value Assessment and 9 Financial Results". The "Process" is described: 10 "The treatment of Value Factors, including the 11 weightings and sensitivity analyses to be applied to the 12 scores in the evaluation is described in paper ... This 13 paper was agreed by the Procurement Board late last year 14 and lodged [with] the programme lawyers prior to receipt 15 of tenders." 16 If we continue, please. If we look at 17 paragraph 7.2, please, it is on page 17. Thank you: 18 "The scores resulting from the assessment ... are 19 shown below. The layout reflects the pre-tender 20 agreement that the factors would be categorised as 21 either 'Characteristics' or 'Viability'. Viability 22 would consider the soundness of the essential services 23 in terms of the internals of the service delivery, while 24 Characteristics would look more at the external factors 25 affecting the potential success of the services." 79 1 You didn't give these scores; is that right? 2 A. No, no, no. These scores came from an overall scoring 3 process which the whole team produced and then they went 4 through a process of weightings, et cetera, to actively 5 ratchet up to this level. 6 Q. If we just quickly look down at the three service 7 providers there, we can see, in relation to the ten 8 value factors that we looked at, do you remember in 9 annex C, the first of them customer acceptability, 10 Pathway is joint lowest; Flexibility, Pathway is the 11 middle; reliability and support, Pathway is the lowest; 12 innovation, it is the middle of the three; as with 13 staff/agent acceptability. 14 Then "Viability": fraud free method of payment, 15 Pathway is the lowest; credibility of delivery, Pathway 16 is the lowest; management capability, Pathway is the 17 lowest; start-up, Pathway is the lowest; stability and 18 coherence, Pathway is the lowest. 19 Is that right? You remember that at the time, do 20 you? 21 A. I don't remember the exact scores but I remember that 22 shape of things, yes. 23 Q. Then over the page, please. We will skip over that 24 figure and look at the table where weightings are 25 applied to the scores that have been given. We can see 80 1 that under "Characteristics", Pathway comes bottom and 2 under "Viability" Pathway becomes bottom, yes? 3 A. That's what it says, yes. 4 Q. Then can we go to page 21, please. This is analysis and 5 conclusion. There is a list of people there who are 6 said to have considered the results of the evaluation 7 and reached a conclusion. I do not think you are one of 8 them; is that right, that you didn't participate in 9 this? 10 A. Correct, yes. This was considerably higher up the food 11 chain than I was allowed to be. 12 Q. But you contributed data on which this group of people 13 made their decision? 14 A. Yes. The scores that we did and the risks that we 15 produced went into that process and Tony Johnson, who is 16 on that list there, he was, I believe, was managing -- 17 he was representing the Demonstrator Stream. 18 Q. Then if we look at 9.3: 19 "The group considered the results from the various 20 streams of activity ... 21 "the Contracts Assurance review ranked the suppliers 22 in the order Pathway [I think, top] IBM, Cardlink ..." 23 In the last sentence there in (a): 24 "Pathway should be preferred to IBM unless IBM's bid 25 offered a considerable price advantage." 81 1 "the Financial Evaluation ... had shown IBM with the 2 lowest cost of service but Pathway sufficiently close 3 for the two to be regarded as virtually equal ..." 4 Then the value factor, that's the thing we've just 5 looked at: 6 "... a close much between the three suppliers in 7 terms of the 'external' factors ... the order within 8 that being Cardlink, IBM and Pathway [that's Pathway 9 bottom]. On 'internal' factors covering the soundness 10 in terms of service delivery (eg stability and 11 coherence, fraud-free method of payment) the order was 12 again Cardlink, IBM and Pathway, with the first two 13 being significantly ahead of the third." 14 Then over the page, please. If we just go to the 15 conclusion at 9.9. So, in light of what's been said, 16 what we have just read, the group "unanimously" decided 17 that the contract should be awarded to Pathway. If we 18 go back to 9.7, please: 19 "The group recognised that an award to Pathway would 20 imply a need for a proactive management stance by 21 sponsors, notwithstanding the improvement noted by the 22 Contracts Stream since the restructuring immediately 23 prior to [invitation to tender] issue. It would also 24 require sponsor staff to work closely with Pathway on 25 fraud prevention measures, although given the changes on 82 1 fraud risk made by the other two bidders in their 2 re-tenders most of this work was likely to be required 3 whichever supplier were chosen. 4 "Whilst acknowledging the complications of selecting 5 Pathway, the group considered this a far preferrable 6 prospect to the consequences of awarding to IBM (in the 7 unlikely event of their being regarded as PFI-compliant) 8 given IBM's stance on fraud risk transfer and other 9 factors, most notably limited liability." 10 Do you think what the phrase "proactive management 11 stance" means? 12 A. I think it meant what we were trying to do and -- within 13 the assurance team and what we, I have to say, in some 14 way, failed to do, because we weren't allowed to and 15 were blocked doing it. 16 Q. Was it possible to take a proactive management stance 17 under a PFI when the service -- excuse me -- supplier 18 said "Trust me, I'm a doctor, you can't see what's in 19 the black box or how we've designed it"? 20 A. I think it would be very difficult to. When I read 21 this, and I don't know whether I had previously seen 22 this document or not, but it did strike me as ironic 23 that the problems that we had and then, to be honest, 24 suffered for at least two years in the period, sort of 25 1997 to 1999, in trying to do assurance, was kind of, 83 1 I think, flagged up here as "Oh, you're going to need to 2 do more of this". 3 And I think there was a -- my view when I was 4 working on this was -- looking at this end of the 5 statement, is there should have been more support from 6 up on high within BA and POCL to make sure that we were 7 able to do our task and whether that required 8 a different contract or a better contract, something 9 done to actually enable us to manage that the risk was 10 in here. 11 Q. Was it communicated to you that, in the light of the 12 award of the contract to Pathway, there needed to be 13 a proactive management stance in the future? 14 A. I have no recollection of that. 15 Q. By that, do you mean that it may have happened but you 16 now do not remember or more positively -- 17 A. I am sure I would have remembered if there had been 18 a meeting to say "We are going to give it to Pathway 19 but, hey guys, we are going to need to set something 20 special up to manage them". My understanding was we got 21 to the end of the evaluation phase, all this took place 22 and then people transitioned onto the assurance phase. 23 But I don't remember any -- I have got no record of any 24 process, meetings, whatever, that said, "because it is 25 Pathway, we need to do X". 84 1 Q. Thank you. Can we look please at WITN05970100, your 2 witness statement, at page 22, please. It is 3 paragraph 66. You say: 4 "My understanding from reading document ..." 5 You quote it, that is the document I have just shown 6 you: 7 "... is that Pathway were the only one of three 8 [Service Providers] who were considered PFI-compliant 9 regarding fraud risk transfer for BA; the other two 10 [Service Providers] were not PFI-compliant. Ironically 11 of course, BA then withdrew from the project in 12 1998/1999 rendering this factor in the decision rather 13 irrelevant." 14 Now, I think what you are doing there, you are 15 referring to the fact that an important factor in 16 awarding ICL the contract was the non-compliance of the 17 other two service providers, in particular in relation 18 to fraud risk transfer? 19 A. Yes. 20 Q. The irony that you are referring to when you say 21 "ironically of course", would this be right, a twofold 22 irony: firstly, the Benefits Agency then withdrew in 23 1999, so the risk of fraud on the Benefits Agency was no 24 longer a relevant consideration? 25 A. Precisely yes. 85 1 Q. And, secondly, this was no longer a PFI contract in 1999 2 and, therefore, the extent to which either of the two 3 failed bidders were PFI compliant was, in 1999, now 4 irrelevant? 5 A. I hadn't, when I wrote paragraph 66, thought of that 6 aspect, for the simple reason that, by the time we got 7 to 1999, as I said before, the damage was done. We got 8 to the system as it was at that point and deciding this 9 wasn't PFI any more wasn't going to change the way in 10 which Pathway had behaved over the previous three years. 11 Q. I want to ask you about that. Can you recall whether 12 any thought was given in the spring of 1999, perhaps up 13 until May 1999 after the Benefits Agency withdrew and 14 these two points, that I have mentioned, carry no weight 15 as to whether the process of letting the contract to 16 Pathway should be undertaken again, given what we have 17 read about of ICL Pathway's weaknesses on technical 18 risks? 19 A. There were, I believe, many discussions that went on as 20 to whether the removal of themselves by BA meant that 21 the project should stop or go on or how it would go on. 22 These were well above my level of discussion but they 23 were, I believe -- from what I have read, all the way up 24 into government -- discussions as to whether the project 25 should go ahead or not. 86 1 Q. I will put it another way. In the documents that we 2 have read, ICL Pathway at its bid, suffered from, over 3 time, an assessment by POCL of serious and sustained 4 technical risks. Was thought given to resetting -- 5 pressing the reset button to say "Look, you got over the 6 line, Pathway, because of the strength of its 7 contractual provisions, where the other two failed 8 bidders failed? We now need to undertake a fundamental 9 re-assessment, given what we know, albeit some of it is 10 obscured by the black box, of the technical risks of the 11 ICL Pathway programme"? 12 A. I'm not aware that there were discussions of reassessing 13 the Pathway solution at that point. There were 14 certainly discussions as to whether the project should 15 go ahead or not. But bearing in mind at this point 16 whilst the Benefits Agency were extracting themselves, 17 POCL and Pathway were heading towards trying to get 18 towards acceptance and trialling, et cetera. 19 The efforts that were going on on the project, at 20 that point, I think, were primarily towards that rather 21 than a fundamental reappraisal. I believe that the Post 22 Office missed a trick by not modifying the contract in 23 some way that forced us to be able to get hold of 24 everything that we hadn't been able to before. 25 Q. We'll come on to the possible missing of the trick 87 1 a little later, possibly, this afternoon. Can I turn to 2 your fourth role then. That's your involvement in the 3 development and assurance of the Horizon system after 4 the award of the contract to Pathway in April and 5 May 1996. 6 You explained in your witness statement, is this 7 a fair summary, that this part of the project was, 8 within POCL, arranged in a similar manner to the 9 demonstrator phase, with teams being allocated to 10 specific areas within the project and you were allocated 11 the POCL infrastructure aspect, just as before? 12 A. Correct. I'm sorry, the one difference is, at that 13 point, is -- at that point, the support services part, 14 I think there was a separate group who were going to 15 deal with implementation and rollout, and I think they 16 took on issues -- well, consideration of anything to do 17 with the rollout side. 18 So, at that point, POCL infrastructure was the 19 technical side of the infrastructure rather than 20 including the support services. 21 Q. Thank you. Can we turn up WITN05970100. This is your 22 witness statement. Can we turn to page 24, please. If 23 we just look at paragraph 72 and 73. You say: 24 "As [you] remember it, [your] role in assurance was 25 to try to monitor the development of the service to gain 88 1 increasing confidence in the emerging Pathway solution, 2 to assure its performance/security, whilst also trying 3 to support them by providing access to resources in BA 4 or POCL where needed (admittedly that was more relevant 5 for the Applications teams). 6 "73. In addition to de-risking the project and 7 facilitating the flow of information, we were also 8 mindful of the fact that eventually there would need to 9 be both a contractual acceptance of the 10 solution/services and Release Authorisation decisions, 11 and that gaining confidence and knowledge throughout the 12 process should make this simpler -- basically it was 13 a means of avoiding surprises downstream." 14 In relation to the activities that you describe 15 there, is it right that you set up a technical assurance 16 forum? 17 A. We did. We tried a number of different approaches. The 18 technical assurance forum was one of them. The 19 intention of the technical assurance forum was to try 20 and give some structure because we had been criticised 21 by -- I think, for wanting access to everything so we 22 tried to have a more structured approach to it. 23 Q. Who was criticising you? 24 A. That was criticism coming from people who were trying to 25 access within Pathway. 89 1 Q. They were saying "You're seeking too much information 2 from us"? 3 A. Yes. 4 Q. You say, elsewhere in your witness statement, that 5 Pathway refused to disclose design documentation to you 6 and that the Pathway management obstructed access to 7 documentation and sometimes cancelled meetings. 8 A. Yes. 9 Q. Did that happen at this stage? 10 A. Yes. 11 Q. Is that why you used the word "obstruction" by 12 ICL Pathway? 13 A. Yes, but more generally in that we found that, at 14 a working level, if we could get contact to individual 15 engineers or individual managers of areas, they would 16 generally be helpful. When we tried to get hold of, for 17 instance, formal copies of documentation or access to -- 18 either it was denied to us -- well, it was denied to us 19 either because it was deemed not appropriate or in some 20 cases -- well, it probably didn't exist. 21 Q. So was the technical assurance forum an attempt to find 22 your way through this obstruction? 23 A. Yes. 24 Q. Can we look, please, at WITN05970138. These are the 25 terms of reference for the technical assurance forum and 90 1 you can see they are dated 9 December 1998. I think 2 this is a document that you co-authored with a Mr Long; 3 is that right? 4 A. Yes. 5 Q. Was he from ICL Pathway? 6 A. He was from ICL Pathway and he was the sort of contact 7 partner I had at that point for this work. 8 Q. I think it is right that the technical assurance forum 9 had already met a number of times before these terms of 10 reference were drawn up because I think we got some 11 agendas for some previous meetings that pre-date 12 9 December 1998. Would that sound about right? 13 A. As I said, we went through a number of different 14 approaches to doing this, so quite possibly yes, we had 15 met beforehand and, as part of that, it was "Okay, we 16 want to formalise this", December 1998 being quite late 17 in the process. So that's why I think this wasn't -- 18 there were other papers, I think as I referred to in the 19 witness statement, of slightly earlier approaches 20 towards this. 21 Q. If we look, at the same time, at WITN05970129, we can 22 see an agenda for a technical assurance forum, it is the 23 second meeting, "Meeting 2", and it is dated 24 30 September 1998, so a couple of months before the 25 terms of reference document. 91 1 But you are, I think, saying there were a number of 2 approaches that you tried to take to find a pathway 3 through this? 4 A. Correct. 5 Q. We can just look at the document on the left-hand side, 6 please. Can we look at page 2 of the document, please, 7 and read the "Background" at paragraph 2: 8 "One of the responsibilities of the Product 9 Assurance Group within the POCL Horizon programme is to 10 make a recommendation to the Release Authorisation Board 11 ... for each release of the Pathway service as to the 12 fitness for purpose of that release for its intended 13 environment. One of the viewpoints for this 14 recommendation is from the Technical Assurance angle, 15 and in particular covering the (as opposed to 16 functionality) characteristics of the service in four 17 key areas, [namely]: 18 "performance and scalability 19 "integrity 20 "service availability and resilience [and] 21 "security." 22 What does "integrity" mean in this context? 23 A. "Integrity" I would take in this case to basically mean 24 that what data comes in is correctly processed and comes 25 out without being modified, without being destroyed, and 92 1 so and so. 2 Q. Reliability and credibility. 3 A. I would say reliability of the service is does it work 4 consistently over time? Integrity is, you know, is 5 it -- it is the sort of word I use all the time, I can't 6 think of a better way of describing it. Integrity: is 7 there a clear trail from the data that goes in to the 8 data that comes out at the other end? Is it immutable, 9 is it reliably going to be processed? Sorry, I have 10 a mental block on the wording around it. 11 Q. That is okay. Thank you. You say in your witness 12 statement, the cross-reference is paragraphs 76(a), (b) 13 and (c) and you cross refer us to documents that you 14 wrote in July '97, January '98 and May '98. I'm not 15 going to ask you about those but, instead, ask you some 16 more general questions. 17 We can take this document down now. You were still 18 operating under a PFI contract at this time? 19 A. Correct. 20 Q. Given you were still operating under a PFI contract, 21 where the supplier was saying that, so long as the 22 outcomes would be delivered in accordance with the 23 requirements of POCL, that POCL need not worry, why did 24 you, nonetheless, engage in this technical assurance 25 process? 93 1 A. The trite answer to that is because it was my job to. 2 The probably more useful answer to that is because, to 3 try and assure the success of the project, we believed 4 we needed visibility of what Pathway were doing and that 5 included not just the functional characteristics but 6 also the technical characteristics -- the how it 7 behaved -- and to understand that, that we wanted and 8 needed visibility of the design and that we needed 9 visibility of things, like how they had analysed 10 (unclear). 11 Q. I think you reference those kind of answers in 12 a document that you wrote back in July 1997. I wonder 13 whether we could look, please, at WITN05970113. This is 14 written by you under the authority of Mr Meagher? 15 A. John Meagher [pronunciation corrected] he is, not 16 "Meeger". 17 Q. Thank you very much. Dated 23, I think, or perhaps 18 22 July 1997. 19 A. Yes. 20 Q. You will see the title of it, "The Level of Information 21 Needed for Technical Assurance -- A Discussion Paper". 22 If we can go forward to page 3, please. Under the 23 heading "The Need for (Technical) Assurance", you set 24 out some of the key reasons for assurance under five 25 bullet points: 94 1 "to minimise [the] risk course to sponsors -- major 2 significant business/political risk are still owned by 3 the Contracting Authorities [that is POCL and BA at that 4 time], and contractual remedies are not sufficient to 5 mitigate the effects of failure of the Pathway service. 6 "to ensure compatibility between various suppliers' 7 solutions -- the 'end to end solution' includes services 8 provided by both Pathway and a variety of other 9 organisations ... 10 "to control solution drift ... " 11 I think a separate paper was written about that: 12 "to enable 'acceptance' -- a number of requirements 13 ... relate to attributes of the service rather than 14 business functionality. 15 "to inform the Release Authorisation Process -- we 16 need to have an adequate understanding of the solution 17 to be able to make an informed decision on the fitness 18 of a Release ..." 19 Do those five points, in the way that I have 20 summarised them at some speed, set out the reasons why 21 you, nonetheless, engaged in a technical assurance 22 process? 23 A. Yes, I think those are good. 24 Q. To what extent was that process necessary as a part of 25 the route to acceptance? 95 1 A. Do you mean by that that, if we had gone away for 2 two years and come back and waited for acceptance 3 specifications and ticked them off? 4 Q. Yes. 5 A. I think we -- point 4 -- we needed to understand the 6 solution to be able to inform the acceptance process. 7 ICL Pathway had to produce acceptance specifications, 8 ie how they were going to meet a number of these 9 requirements. Without knowledge of the solution we 10 couldn't really evaluate those acceptance criteria. 11 If they said requirement 1, 2, 3 was going to be met 12 by it being green, what would that mean if we didn't 13 understand what was underneath it? 14 What we were trying to do here was get to the stage 15 where we understood enough so that acceptance could be 16 done and that acceptance could be done, not just with 17 the functional characteristics, which are maybe fairly 18 easy to do, but with some of these non-functional ones. 19 Q. The third point is how did what you were trying to 20 achieve here sit with ICL Pathway, given the provisions 21 of the PFI contract and their approach to carrying them 22 into effect on the ground? 23 A. Badly, I think is the answer to that. We tried through 24 these various papers and various approaches to Pathway 25 to engage with them, both this one in 1997 and then the 96 1 paper you had at the end of 1998, but I think Pathway's 2 approach was they were head down trying to do the job 3 and I think they viewed us as a distraction. 4 Q. Can we look at your witness statement, please, 5 WITN05970100, at page 27. You say in paragraph 80: 6 "Rather perversely, we did make some progress on the 7 POCL Infrastructure side (I say perversely as you might 8 expect that the deeper into the software stack, the 9 furthest from the application, the more reticent they 10 might be, but the opposite seemed true) with Pathway 11 people providing 'informal' access to a document known 12 as the TED (Technical Environment Description). This 13 technical infrastructure area was making progress and 14 did not expose major problems, which might explain why 15 the people here were more open. We also had more 16 success in the underlying Security areas, as we had 17 a specific requirement which required Pathway to create 18 a Security Functional Specification (SFS)." 19 So there you are pointing out a perversity that you 20 are diving deep into a software stack you might expect 21 resistance but you didn't get it? 22 A. Correct. 23 Q. Can we contrast that with what you say in paragraph 82: 24 "In other areas however, in particular the POCL 25 Applications side, there was explicit refusal to allow 97 1 access to design documentation for much of the time. 2 I suspect this was largely because such documentation 3 did not exist at the time or that it was of such quality 4 that opening it up to us would have exposed Pathway in 5 this area." 6 What do you mean by your reference to the POCL 7 applications side? 8 A. So POCL applications included EPOSS, APS -- the 9 automated payment service -- included BES which was the 10 Benefit Encashment Service, which was part of the 11 overall benefit payment service but the bit that ran on 12 the counter, so POCL applications were those three at 13 the time and I think NFS was added to it. 14 Q. You say that "explicit refusal to allow access" was what 15 we were met with. 16 A. Yes. 17 Q. What was the reason that was given for this refusal to 18 allow access? 19 A. I believe it was the, you know, "not appropriate under 20 the contract". The, what I called, the "PFI card". 21 Q. You say that you suspect that this was essentially 22 a front, is that right, because the document did not 23 exist or was of poor quality? 24 A. That's what I suggest there, yes. 25 Q. On what basis do you hold that belief -- or did you hold 98 1 that belief? 2 A. Partly because it seemed very odd that in some areas we 3 were getting access and was very good in the areas 4 where -- for instance, the document for TED, the 5 applications, we had more resistance. I say this, 6 I suspect is largely because what we have learnt since, 7 in particular around EPOSS, and what I think is backed 8 up by certain documents that the Inquiry gave me 9 yesterday in appendix E, that says precisely this. 10 Q. That the documentation didn't exist or that it was of 11 low quality? 12 A. Yes, what's in the EPOSS task force document and the CSR 13 audit, documents the Inquiry showed me yesterday seemed 14 to say precisely this. 15 Q. We will come to that this afternoon, the EPOSS, but to 16 summarise what you are saying, you have now been shown 17 some internal ICL Pathway analysis of difficulties that 18 they were encountering with the EPOSS system; is that 19 right? 20 A. Yes. But it does mention in those documents the absence 21 of design documentation and the fact that some 22 documentation had to be reverse engineered from what had 23 been written after the event, therefore wouldn't have 24 existed at the time that you are referring to here. 25 Q. Can I turn to the effect of what all of this meant, the 99 1 consequences and look at paragraph 84 of your witness 2 statement please which is over the page. You say: 3 "The result was that, except in areas where we had 4 an explicit right in the Contract to a document ... we 5 only had limited or partial visibility of the emerging 6 Pathway systems, or of their design/developmental 7 approach. This meant that we could not gain confidence 8 of what Pathway were creating (or its suitability or 9 fitness for purpose) or have confidence in how Pathway 10 were developing (and therefore have quality mechanisms 11 were in place)." 12 Summarising, you say you couldn't see what they were 13 creating and you couldn't, therefore, have confidence in 14 how they were developing the system; is that right? 15 A. Correct. 16 Q. You obviously knew that that was a problem at the time; 17 is that right? 18 A. We suspected that the -- we knew that the lack of 19 visibility we were getting was impeding our work on 20 assurance. We obviously didn't know what kind of 21 outcome that was going to be. If you take the view that 22 it was Pathway's responsibility and it was their 23 responsibility to have the right controls in place, you 24 could say the effect of all that would be zero because 25 they should have got on and done it. What we came out 100 1 with though, at the end of the process in 1999, was we 2 didn't have evidence to have confidence in the system at 3 that point. 4 Q. Just before the break then. You, I think, recorded some 5 of this at the time. Can we look, please, at 6 WITN05970118. 7 I think this is a document you wrote at the time; is 8 that right? 9 A. Yes. 10 Q. In your statement you tell us that it dates from 1998. 11 Can you be anymore precise than that? 12 A. I can't at the moment without -- unless it is within 13 the -- can I just check what WITN number this is? 14 Q. 118. 15 A. According to page 2 it was 1 October 1998. 16 Q. I saw that in the index. Where did that date come from? 17 A. That I believe was the date on the file. 18 Q. So the properties of the document? 19 A. Yes. 20 Q. Thank you. Who was this addressed to? 21 A. I believe this was a document I produced for discussion 22 within the product assurance team. 23 Q. So who would it have gone to within the product 24 assurance team? 25 A. The team was led by John Meagher and there were other 101 1 members of the team and I presume it would have been 2 circulated within that team. 3 Q. Do you know whether the document was ever escalated 4 above Mr Meagher? 5 A. I presume it was discussed but I don't know. 6 Q. Looking at the first section, "Assurance vs issues 7 management", you say: 8 "Our role is assurance, but up to now we've had very 9 little opportunity to do more than manage issues as they 10 come up. In the absence of Pathway's co-operation on 11 assurance (across the board), and the ability to engage 12 in proper review activities -- the V&V ..." 13 What do you mean by V&V? 14 A. Validation and verification. 15 Q. "... we are not going to be able to provide the 16 'assurance' that sponsors expect. 17 "We end up giving Pathway false credibility and 18 doing ourselves no favours if we give the impression of 19 assurance. Pathway see little benefit in overall 20 assurance, they choose to involve us when it suits them 21 but consistently fail to provide long-term access. If 22 they refuse access, what happens -- do we have any 23 leverage? In effect we have no clout -- what does 'no 24 assurance' mean in the end? 25 "So either we have to achieve the coverage/depth 102 1 that we require to provide a credible level of 2 assurance, or we stop pretending we can do it, and 3 revert to 'issues management' as our prime role. 4 "[A] decent level of assurance is only going to come 5 from high level (Dave Miller and above) pressure on 6 Pathway, as a condition of signing up to 7 yet-another-plan, as a risk-minimisation action -- the 8 little local agreement approach is doomed." 9 Does that accurately reflect how you felt at the 10 time as part of the assurance process? 11 A. I think it does yes. And the comment at the end, "the 12 local agreement approach", I think, by that, I meant us 13 trying to talk nicely to -- I think it probably would 14 have been Terry Austin or whatever at that point -- 15 wasn't really getting anywhere. That's the only way we 16 were going to move this forward is if we had, actually, 17 support from the top of the programme. 18 Q. That is referred to in some of the documentation and in 19 your witness statement as "informal access to 20 information" and "informal access to some documents"; is 21 that right? The little local agreement? 22 A. In some cases, and in particular on the more technical 23 side, we did get, for instance, informal access to this 24 document called the TED. What I would have wanted at 25 that point was a controlled copy of the TED, so if it 103 1 got updated that we saw those updates. There is not 2 much point in assuring a document and it changing 3 six months later. 4 What the little local agreement approach is saying 5 is, that for some documents we just did not get access 6 and that continuing to apply pressure doesn't -- wasn't 7 going to give us what we needed at that level. 8 MR BEER: After lunch can we turn to what the little local 9 agreement and the informal access was showing you. 10 Is that an appropriate moment, sir? 11 SIR WYN WILLIAMS: 2.10 pm all right with everyone? 12 MR BEER: Yes. Thank you very much, sir. You should give 13 Mr Folkes the warning. 14 SIR WYN WILLIAMS: I am sure you are aware of it but you 15 shouldn't discuss your evidence with anyone over lunch. 16 I imagine it is the last thing you want to do anyway. 17 A. Correct. 18 SIR WYN WILLIAMS: I will see you at 2.10 pm. 19 A. Thank you. 20 (1.08 pm) 21 (The short adjournment) 22 (2.07 pm) 23 MR BEER: Good afternoon, Mr Folkes. We were, before lunch, 24 looking at what you describe in your witness statement 25 and in some of the contemporaneous documents as informal 104 1 access to some information and some documents that ICL 2 Pathway allowed you to see. 3 Can I look at some examples of that please. If we 4 turn up, please, WITN05970143. 5 If we just look at the top of the page, we can see 6 that this is a fortnightly report from you to 7 Mr Meagher; is that right? 8 A. Yes. 9 Q. Did you produce these on a fortnightly basis? 10 A. Yes, or thereabouts. 11 Q. What was the purpose of them? 12 A. John was leading the product assurance area, so it was 13 fortnightly reporting of progress and issues. At this 14 point, I was doing work for two people, for John and for 15 Gareth Lewis, who headed up the fault and security group 16 on the programme for BA, and so I was splitting my time 17 between the two. So it was also just trying to justify, 18 if you like, what time I was spending for each of the 19 two bosses I had. 20 Q. At this stage, what was Mr Meagher's position? 21 A. I think the group at that point was called design 22 assurance, and so he was, I think, manager of design 23 assurance, it may not be quite the right name but that 24 was the gist of his role. 25 Q. Was he next up the chain from you? 105 1 A. Yes. 2 Q. Do you know what happened to these reports? 3 A. I presume that key issues from them were then fed 4 further up the chain. 5 Q. By "further up the chain", where do you mean? 6 A. I presume, I'm just trying to work out what the 7 structure would have been at that point, but to either 8 the programme director or the deputy programme director 9 at that point. 10 Q. So Mr Miller? 11 A. At that point in 1998 may have still been a Benefits 12 Agency -- Peter Crahan, I think, was the programme 13 director at that point. 14 Q. Could we look at the foot of the page, please, under 15 paragraph 3. You say: 16 "Current Issues and Concerns 17 "Testing/Acceptance: documentation from PDA Testing 18 at Borough ..." 19 Does that mean Borough, the place in London? 20 A. Okay, the Post Office had a site, I think it was like 21 an industrial unit or warehouse, that was used for test 22 equipment and that was at Borough. 23 Q. Ie the place in London? 24 A. Yes. 25 Q. "... re Model Office and End to End ..." 106 1 Can you explain briefly what "Model Office" and what 2 "End to End" are? 3 A. These were two different test phases or test activities. 4 "Model Office" was actually trying to operate a post 5 office, presumably you would have had two or three 6 terminals and tried to do real transactions and look at 7 the processing within the operating environment. "End 8 to End" would have been trying to look at the process 9 from transactions being performed in the office and then 10 feeding all the way up through the central systems and 11 onwards to, I presume, TIP or to the relevant Benefits 12 Agency systems. 13 Q. But: 14 "... still steadfastly ignoring anything to do with 15 Acceptance. Hopefully will be some change of direction 16 following the 'week long workshop' this week, given 17 Mary's involvement." 18 Who was Mary? 19 A. Mary was a -- I can't remember her surname -- but 20 I believe she was a contractor, probably from PA, who 21 was working for the programme and I think she was 22 working at testing. What this is, is actually probably 23 more of a, sort of, internal issue, regarding -- those 24 two test scripts done at Borough were trying to put 25 together tests. 107 1 My concern was these may have been valid tests to 2 test the functionality of the system but they weren't 3 tied in, in any way, with what was going to need to 4 happen with regards to acceptance. Given the big hurdle 5 we knew was going to happen -- obviously at this point 6 we thought it was going to happen sooner than 1999 -- 7 but the big hurdle that the programme needed to get 8 through would be acceptance, we were looking at how 9 testing and acceptance would work together. 10 Q. Thank you. The next bullet point: 11 "Release 2 and Acceptance: Given the nature of the 12 hangouts for Nile 2, a significant number of acceptance 13 criteria (eg for security) are likely to be 14 'unacceptable' until a subsequent release. Are we 15 content that we understand the contractual and other 16 implications (eg re OP Trial, Rollout, etc) of 'partial 17 acceptance'." 18 Can you explain to us what you meant by that bullet 19 point? 20 A. Nile 2 was the name of release, in the same way that we 21 had releases called Congo, which I think were -- the 22 initial trial was a release called Congo. I am not 23 quite sure what the release called Nile was but it was 24 obviously the release that was coming up at about that 25 time of a wider system. The concern there was that 108 1 a number of the acceptance criteria that we had around 2 security, weren't actually going to be satisfied by Nile 3 or Nile release 2, because Nile release 2 had been pared 4 down. 5 Therefore, the question I'm asking is, if we go 6 ahead and do Nile 2, that may be a sensible process to 7 go through to iteratively get some functionality going 8 but, if it didn't prove the acceptance criteria, we 9 would not be able to give them acceptance. So we needed 10 to, as a programme, have a joined-up approach as to 11 whether we were giving them partial acceptance, whether 12 we were giving some sort of acceptance that it needed to 13 be then rerun or, for the purposes of Nile, we were not 14 giving acceptance at all. 15 Q. I understand. Then if we scroll down there are three 16 bullet points that are all to do with EPOSS. I want to 17 concentrate on those, please. 18 The first of them: 19 "EPOSS -- application: Evidence emerging from the 20 EPOSS workshops that the emerging product is likely to 21 be non-conformant in a number of areas, and will miss 22 the (possibly unwritten) business rules in a large 23 number of areas. The product does not appear to be at 24 the stage one would expect given the closeness to its 25 entry to a testing phase -- Pathway admitted to several 109 1 'holes' where they don't yet have a solution (eg Cash 2 Account)." 3 Just dealing with that first bullet point. How 4 serious an issue was that? 5 A. Well, I would view that extremely serious because EPOSS 6 is one of the core parts of the system. As background 7 to this, at about this time we did attend a series of 8 workshops that were put together by the POCL 9 applications team with Pathway. I attended those from 10 a technical perspective. These were workshops to take 11 the programme people through the emerging EPOSS 12 application. This was kind of being done, "Given we 13 didn't get any documentation at this point, then show us 14 the thing and let us work through it". 15 It is obviously not a particularly good way of 16 reviewing it but it was an important thing to do. We 17 went through it and found there was some functionality 18 that the relevant experts we got in there from POCL 19 basically thought was non-conformant and, in some cases, 20 didn't appear to meet the business rules they were 21 expecting. 22 I think in the possibly unwritten bit I put in there 23 is there was an issue with the detail of the 24 requirements, partly, I think, because of the whole 25 issue we talked about before lunch, about the level of 110 1 requirements. POCL did put in experts to try and help 2 Pathway with some of the business functionality around 3 EPOSS, in particular. So they may have helped them in 4 sort of joint working but they may not necessarily have 5 been written down business rules at that point. 6 Q. What does the reference to the "cash account" mean? 7 A. Okay. So, the core -- the key output that every office 8 produced every week was a document called the cash 9 account. It was a fairly complex document every office 10 had to produce. I have got one here if you want to see 11 it but it was a key financial report that went off to 12 the centre, got keyed by Chesterfield and then was used 13 for a whole series of purposes by the centre. 14 EPOSS was meant to effectively replace the cash 15 account by initially electronically producing the cash 16 account. I think, originally, it was going to 17 electronically produce and feed to TIP but also print 18 for signature and then, eventually, they would dispense 19 with the paper copy. 20 So the cash account was the key financial report 21 coming out of the office. It had a whole series of 22 steps that were needed to create the cash account that 23 was done manually that every -- you know, 20,000 24 subpostmasters knew how to do manually, and this had to 25 be replicated by EPOSS. 111 1 Q. You were saying that Pathway were admitting that they 2 didn't have a solution to the creation of the 3 replacement for the cash account document? 4 A. Yes. 5 Q. Was that issue resolved subsequently to your 6 satisfaction? 7 A. It was an issue that was resolved in that, yes, they did 8 move on and were able to produce a cash account and the 9 system couldn't have gone live if we couldn't. To my 10 satisfaction? Because this was in the application area, 11 I was sort of trying to pull together the things we were 12 finding. That wasn't actually one that I was 13 responsible for chasing through. So I can't say I went 14 back and spent the next six months working on the cash 15 account. 16 Yes, I believe it did get resolved but what was of 17 concern to us, at the date of this document, which 18 was -- 19 Q. January 1998. 20 A. Which was meant to be fairly near to when we were 21 getting into detailed testing, but not be able to create 22 the core accounting document out of the system, at that 23 point, didn't bode well. 24 Q. The next bullet point on EPOSS: 25 "EPOSS -- design approach: Very concerned about 112 1 Pathway's (apparent) design approach for EPOSS, which is 2 totally inappropriate for an application of this 3 complexity -- appears to be based on reverse engineering 4 a product which has been cobbled together first by 5 someone who is no longer with Pathway (and left little 6 documentation) and since by Escher. This is a very 7 dangerous approach for a product of this nature and 8 importance, and I do not believe that the risk can be 9 adequately mitigated through testing alone." 10 First of all, some of the words that you use there, 11 "very concerned", "totally inappropriate" and "very 12 dangerous"; that's quite strong language. 13 A. It is. 14 Q. Why did you use such strong language? 15 A. So EPOSS was the core part of the system that was 16 providing the accounting, providing the creation of the 17 cash account and everything that went before it in that 18 process, and was the key document where a subpostmaster 19 or branch manager would report to the centre what they 20 had. So it was fundamental to POCL operating and being 21 able to account for the business and the money in its 22 network. 23 EPOSS is fairly complicated. I saw Mr Cipione's 24 report and I felt is kind of underplayed a bit the 25 complexity. It is more than just a POS system. Because 113 1 of the sheer complexity, there are 170/180 different 2 products that you could transact -- different product 3 types you could transact at a post office, each one has 4 subtly different ways of being accounted for, the cash 5 account form is complicated. 6 It is a complex piece of software and, in my view, 7 it needed to be properly analysed and designed. It 8 wasn't something that should be just, sort of, put 9 together from the user interface side or -- I see the 10 use of RAD, rapid application development, was talked 11 about. There were some bits of software which are 12 suitable for that kind of approach and others which 13 aren't. 14 Q. Which was this? 15 A. Totally unsuitable for the -- emotive language, "cobbled 16 together" -- but totally unsuitable to try and put it 17 together and write up design documentation afterwards. 18 Can I just say, I had, in a previous life, been 19 technical manager on the ECCO projects, which was to 20 create the -- effectively, the EPOSS project which was 21 used in branch offices for a number of years. So I was 22 aware -- I'm talking with some knowledge of the 23 complexity of actually trying to automate that process. 24 Q. Is what you are describing in this bullet point the kind 25 of run-of-the-mill issue that one might reasonably 114 1 expect to see in the development of any large IT system? 2 A. I would certainly hope not, because I would hope that 3 anybody developing a large IT system would use the 4 appropriate design approach for the complexity and 5 importance of the piece of software they are producing. 6 Q. Are you raising here a fundamental issue about the 7 competence and ability of your contractors? 8 A. I probably can't answer no to that. I think I am, yes. 9 Q. What was done about it? 10 A. There were various approaches that we -- that were taken 11 within the programme, as far as joint working with them, 12 as far as trying to get hold of design documentation, 13 et cetera. I think our concern was that we were finding 14 this stuff out, finding out the stage of the product, by 15 doing this review very late in the day, not through 16 reading their own design documentation. Some of what we 17 were finding out here was from, if you like, rumour or 18 what people said in unguarded moments, rather than by 19 official channels. 20 But, you know, this was reported up. I think it is 21 now -- seems to be backed up by other evidence that 22 I have seen in the past 24 hours that I hadn't seen at 23 the time. I don't personally know what pressure was 24 applied to Pathway or what was taken forward to Pathway 25 at this point. 115 1 Q. When you refer to having seen documents backing this up 2 in the past 24 hours, are you referring in particular -- 3 we will come to the document in a moment -- to 4 ICL Pathway's internal report on the EPOSS PinICL Task 5 Force? 6 A. Yes, and also on the CSR+ audit, which echos some of the 7 findings from it. 8 Q. You refer in this document, in the second bullet point, 9 to the application appearing to be based on reverse 10 engineering. You say "appears to be based"; on what was 11 that founded? 12 A. I think it was founded on formal questioning of the team 13 during these workshops. So these workshops were held, 14 I was one of the people attending these workshops. We 15 had obviously up to then trying to get hold of design 16 documentation. We had been denied it possibly because 17 it didn't exist. We then went into these meetings. 18 There were questions that we would have asked through 19 these two or three workshops and we would have been 20 asking them these questions of some of the staff who 21 were there on the ground, who would have given us 22 possibly less guarded answers than management. 23 Q. What's the problem with reverse engineering this 24 product? 25 A. Well, what you have got here is a complex requirement. 116 1 It's got strong requirements for integrity and 2 reliability, obviously because it is handling the 3 accounting in the office. It's got to be able to 4 function reliably under hostile conditions, et cetera. 5 What we found out, I think, by talking to people 6 there, was that they had started putting a product 7 together, they had had -- it was unsuccessful. We knew, 8 I think, that one of their developers who worked on it 9 had left. They then tried to do further work on it. 10 They shipped it over to Escher in Boston to, we were 11 told, complete it. I think "complete" probably meant 12 a lot of rewriting. It had then come back again and 13 then, from what it says here, it appears that then they 14 tried to do more major work on it, reverse engineering 15 what they had already got. 16 It is just not the way that you would want/expect 17 a product that is this mission critical to be produced. 18 Q. Can we turn to the third bullet point, please: 19 "EPOSS -- failure conditions: Significant concerns 20 re operation of EPOSS -- and the office platform in 21 general -- during 'everyday' failure conditions, such as 22 loss of a terminal or LAN ..." 23 Local Area Network? 24 A. Yes. 25 Q. "... connectivity, but similar issues likely to emerge 117 1 in non-failure conditions with shared stock units. 2 Pathway's problem is basically that Riposte gives high 3 integrity for data held on a 'per terminal' basis -- 4 whereas the business requirement is for data to be 5 accounted for on a 'per SU ..." 6 Stock Unit? 7 A. Yes. 8 Q. "... basis; they need to build the integrity for the 9 latter using the facilities provided by the former. 10 Trying to meet this need without a rigorous design 11 method, and without proper failure analysis, is unlikely 12 to succeed. Pathway appear not to understand the 13 business impacts of failure of the accounting process 14 (as opposed to failure of a transaction) and appear to 15 want to rely on the 'it's not going to happen' 16 philosophy (Same may be true of other applications, 17 given Release 1c experience, but have less visibility)." 18 Can you explain what you meant by the distinction 19 that's drawn in this paragraph between a per terminal 20 and a per stock unit accounting approach? 21 A. Okay. So, per terminal, each counter position in the 22 office had its own counter terminal, which has its own, 23 in Pathway's solution, software with Riposte, a message 24 server, which -- message storage -- holds the data. 25 That's fairly easy. 118 1 However, what happened in post offices is post 2 offices are organised on a per stock unit basis where 3 "stock unit" generally meant the till drawer. The best 4 way of thinking of a stock unit is a till drawer. So, 5 if you were working in a post office and you were 6 working on position 1 in the morning, you would go and 7 get your till drawer out of the safe, go to position 1, 8 work on it there. That's what you are responsible for. 9 Your stock unit. 10 If, in the afternoon, you are in position 3, you 11 take the same drawer to position 3 and work on it there. 12 At the end of the day or the end of the week, you want 13 to account for and balance your stock unit and some of 14 your stock units would have been done on terminal 1, 15 some on terminal 3, some somewhere else. 16 So the challenge here was to come up with an ability 17 to reliably -- and there may be failures, for instance 18 the network, et cetera -- come up with a reliable way in 19 which you could actually balance stock units and account 20 for stock unit data where somebody has moved around 21 within the office. 22 If everybody sat on the same terminal all the time, 23 it would be much easier. But they don't because post 24 offices need the flexibility, multiple positional 25 offices need the flexibility of staff to be able to move 119 1 around. 2 Q. Thank you. You say: 3 "Pathway appear not to understand the business 4 impact of failure of the accounting process ... and 5 appear to want to rely on the 'it's not going to happen' 6 philosophy." 7 You say in your witness statement that, by this, you 8 meant that ICL Pathway didn't demonstrate 9 an understanding of the importance of the accounting 10 role of EPOSS, including the balancing of cash and stock 11 in each Post Office; is that right? 12 A. That's what I meant, yes. If I can just say, the 13 distinction here is, when we spoke to them they 14 understood the problems if a single transaction failed. 15 So if you were in the middle of paying a bill and that 16 bill -- say the power went off or something went wrong 17 during that particular transaction, they were -- seemed 18 to be understanding that they needed to tidy up that 19 situation. But less concerned around failure of the 20 underlying accounting. 21 For instance if somebody had, say, worked on this 22 position then that position then this position and then 23 something had gone down. When we tried to challenge 24 them, the impression that I got, what I wrote here, is 25 they were keener to do the, "it's not going to happen", 120 1 rather than to provide us with an explanation of how it 2 was going to be solved. 3 Q. In writing this had you got in mind any difficulties 4 being caused for subpostmasters being able, accurately, 5 to account for their stock and cash at the end of 6 an accounting period? 7 A. Well, yes because that's the whole purpose of EPOSS is 8 for subpostmasters and managers to be able to account 9 for the business and then report that back to the 10 centre. 11 Q. So in a sense describing one of the very issues that 12 this Inquiry is inquiring into? 13 A. Yes, at this point Pathway did not demonstrate, to us, 14 an understanding of this kind of failure condition or 15 how they were going to solve it. Other people may have 16 solved it but certainly, to us as part of this piece of 17 work, they hadn't demonstrated it to us. I don't know 18 whether this particular example is one of the problems 19 which has hit downstream or not but the point I'm making 20 here is they didn't -- we wanted to be able to have 21 proof that they had considered these failure conditions 22 and -- because we knew that such failure conditions were 23 going to happen. 24 Q. You refer to a preference to relying on the "it's not 25 going to happen" philosophy; how was that expressed to 121 1 you? 2 A. I can't remember the exact words but it would be -- 3 I think what I meant by that was -- if we said "what 4 happens if, for instance, the network breaks into two 5 halves?" Say you have a large office, depending on how 6 the local area network in the office is configured, you 7 could end up with two half networks. That depends on 8 how the networks are put together. That may not happen 9 very often, it may only happen in a large number of 10 offices but it still could happen. Therefore, we wanted 11 to ensure that they had taken these things into account 12 or shown us why they weren't going to happen, what we 13 were seeing here -- and it mirrors, to a certain extent, 14 what we found during the demonstrator that they were -- 15 rather than showing they had addressed it they preferred 16 to deny it was going to happen. 17 Q. You expressed a view there that you wanted your concerns 18 listened to and acted upon by ICL Pathway. In fact, at 19 this time, was there a proposal that contractual 20 acceptance be changed and tied to authorisation being 21 given for NR2, new release 2? 22 A. Yes, I believe there was. 23 Q. Can you just explain to us what "contractual acceptance" 24 meant? 25 A. Contractual acceptance was important because it 122 1 basically would have said we were happy with the system 2 and I think it was important to Pathway because it 3 unlocked a major payment. I can't quite remember what 4 it was. But it was important from a financial point of 5 view. Maybe at this point, because it was still PFI, it 6 didn't do that. Acceptance was basically ... 7 Q. I think there were a series of stage payments? 8 A. It was important for Fujitsu to be able to achieve 9 acceptance and then acceptance would then be able to 10 initiate the rollout and obviously then the rollout of 11 the system and that moved us to the next stage and 12 income stream. 13 What there appeared to be -- 14 Q. Sorry to interrupt you. This document can come down. 15 A. What I believe you are referring to, if there was 16 a proposal that rather than having a formal acceptance 17 activity, that they would tie it, acceptance, to the 18 authorisation of a specific release which I think was 19 then called NR2. 20 Q. What was NR2? 21 A. It was another -- it wasn't a full -- it wasn't the 22 final release. I guess the key point here is, if you 23 wanted to accept you might want to accept on the final 24 thing you had and the proposal that came forward was 25 that we would actually -- acceptance would be tied to 123 1 the release authorisation of new release 2, which was 2 a not-final release. 3 Q. I think you wrote a note setting out your opposition to 4 that or outlining the dangers of ICL Pathway's proposal 5 to tie contractual acceptance and the payment of money 6 to them, to a release of NR2? 7 A. I did. 8 Q. Can we look at that, please, at WITN05970119. 9 This is a single page document authored by you. You 10 say in your witness statement that you thought ICL 11 Pathway's proposal was not a good idea and then you 12 reduce your thoughts to writing. This was in 1998. Can 13 you again assist us with the date on this. 14 A. Have I dated it at the bottom of the -- 15 Q. No, it is a single-page document. If we scroll to the 16 bottom you can't see a date. 17 Document 119? 18 A. Acceptance. It seems to be 12 October 1998. 19 Q. Thank you. Again, was that taken from the properties? 20 A. That was taken from the properties, yes. 21 Q. Thank you. If we just read it. You wrote: 22 "A suggestion has emerged that Contractual 23 Acceptance could be given at New Release 2 Release 24 Authorisation, rather than waiting [for] the completion 25 of the formal acceptance process. This note seeks to 124 1 outline likely dangers in this approach." 2 In the first bullet point you say: 3 "The only contractual right that the sponsors have 4 to obtain any assurance that the ICL Pathway service 5 will [meet] the contract is via the Acceptance process. 6 Assurance [without] acceptance is not supported by the 7 contract and has, to date, been only of limited 8 effectiveness due to the reluctance of ICL Pathway to 9 provide access to the detail of their solution. This 10 situation has only been sustainable on the basis that 11 the Acceptance process would provide a backstop for 12 assurance. 13 Then skipping over the next one. 14 "Although a number of the tests used in the 15 Acceptance trials may already have been run in some form 16 ... many of the Acceptance Reviews will not have taken 17 place before the Release Authorisation Board has sat. 18 A large number of deliverables cited for Acceptance 19 Review have yet been made available to Horizon, as these 20 are scheduled to be produced during the Operational 21 Trial. 22 Then: 23 "The current approach will provide assurance to the 24 Release Authorisation Board that the associated 25 'functionality' will work however it will not prove the 125 1 service deliverability." 2 Then: 3 "Conclusion. The current Acceptance Process acts as 4 a safety net for the Contracting Authorities, offering 5 a level of protection from having to accept and rollout 6 an inadequate service. We believe it would be very 7 dangerous to accept any proposal which would remove the 8 protection offered by the Acceptance process." 9 Can you help us who this document was addressed to 10 and seen by? 11 A. I obviously wrote it in a hurry because there are 12 a couple of typos in it. I can't remember who I was 13 asked to produce it for. But the same document has been 14 disclosed back to me by the Inquiry with a handwritten 15 draft in the top right-hand corner so I know it reached 16 somewhere up in the echelons of the PDA as it would have 17 been at that point. 18 Q. The programme delivery authority? 19 A. Yes. I presume it would have gone into the contracts 20 team or those who were negotiating around the contract 21 with Pathway at that point. I presume it would have 22 been at the level of the programme director or 23 immediately below. 24 Q. Can you recall what happened as a result of submitting 25 this paper? 126 1 A. I'm a little confused when I go and look at what was 2 then agreed. 3 Q. Ie, in the acceptance board minutes? 4 A. Yes, as to how it came out. I believe this paper 5 reached the right people but from what I understood the 6 toned down version -- the acceptance board minutes may 7 have been -- the process may have been changed contrary 8 to this advice. 9 Q. Can I turn to a new topic then please. The Project 10 Mentors' report, can we turn up page 39 of your witness 11 statement please. WITN05970100 at page 39. You say 12 in 114: 13 "There was a review performed of the programme by 14 Bird & Bird ... " 15 I'm going to skip what's in brackets given your 16 correction earlier: 17 "... and a consultancy called 'Project Mentors' in 18 March 1998, for which I believe I was interviewed." 19 You refer to an exhibit which you exhibit as your 20 32nd document. 21 Then in 115, you say: 22 "I can't remember seeing the report at the time it 23 was produced, but was invited to comment on it by POCL 24 management in January 1999 and largely supported its 25 findings. My notes from reviewing the report are 127 1 presented as ... " 2 Then you describe them and exhibit those as your 3 33rd exhibit. 4 Can we just look at your note to start with then, 5 please. WITN95970121, thank you. 6 If you just look at your notes. They are said to be 7 prepared in contemplation of litigation at the top 8 right? 9 A. I put that on there just because the document I had been 10 sent had that on and I -- 11 Q. I was going to ask you why? 12 A. I was probably being lazy with the same marking, as it 13 were. 14 Q. If we scroll down, I don't think there is a date on them 15 but I think in your statement you date them as being 16 from January 1999? 17 A. Yes. 18 Q. Then if we just scroll back up again please at the notes 19 and look at the first bullet point, there is a quotation 20 in italics: 21 "'Effective business requirement analysis is 22 required to achieve ...'" 23 Was that a quote from the report? 24 A. Yes. 25 Q. And then you are commenting on it underneath? 128 1 A. Yes. 2 Q. We can see the way that it continues in the second and 3 third bullet point with an italicised quote and then 4 some comments by you. 5 None of those italicised comments appear in the 6 March 98-version of the Project Mentors' report. 7 I wonder whether we could instead look, at the same time 8 as looking at this document, at POL 00038829 and then 9 skip to page 9. 10 If you just scroll down a little bit. We can see 11 this is a 17 December 1998 version of the Project 12 Mentors' report. Then if we scroll forwards please to 13 page 11. 14 Then scroll to 1.3 on the right-hand side. Can you 15 see a sentence under "Scope": 16 "Effective business requirements analysis is needed 17 to achieve a satisfactory, comprehensive business 18 design. This can then be used as the basis for ..." 19 Now, give or take a word or two, that matches what 20 you have written on the left-hand side? 21 A. My apologies, I may be referencing the wrong document. 22 Q. I wanted to establish whether that was so. Drawing 23 those threads together, it looks like your notes on the 24 left are a commentary on or a response to the 25 December 1998 Project Mentors' report, rather than the 129 1 earlier March 1998 version; would that appear to be 2 correct? 3 A. It does appear like that, yes. I don't think I was 4 necessarily aware there were two Project Mentors' 5 reports. I seem to have got them muddled up. 6 Q. Can we get to the substance of what Project Mentors said 7 then, and go to page 13 in the document on the 8 right-hand side, and look at the conclusions. We can 9 lose the document on the left-hand side to make it 10 easier to read please: 11 "It must be remembered that so far we have only 12 performed the requirements analysis for BPS ..." 13 BPS is? 14 A. Benefit Payment Service, which was the overall name for 15 both the BA and the POCL parts of the benefit payment. 16 Q. "... which is predominantly a BA system element. 17 However, from our analysis we conclude that Pathway made 18 no attempt to undertake requirements analysis in 19 accordance with normal industry practice. This [is] 20 despite their having access to the SSR and subsequent 21 requirements since April 1996. Much of this work could, 22 and should, have been done during the demonstrator 23 period." 24 I think that's a conclusion that you agree with. 25 A. This is referring to the BPS and I didn't have a great 130 1 deal of involvement with the BPS but I think the general 2 point about doing requirements analysis I agree with. 3 The only point I might disagree with is how much of that 4 work they should have done during the demonstrator 5 period because at that point, of course, there was still 6 three service providers and so I think you would only 7 expect them to go to a certain point before award of 8 contract. Having got an award of contract you then 9 expect them to go in heavy to do the main detailed 10 requirement work. 11 Q. Thank you, and then over the page, please. Then, 12 under 2.3.4, "Other Elements of the System", they say: 13 "While we have so far only completed work on the BPS 14 system element, we have grave concerns that the same 15 lack of professional analysis will be apparent in other 16 areas as we come to review them. This concern is 17 supported by a number of interviews with Authorities' 18 staff, from which it is apparent that Pathway are loath 19 to release design documents to BA/POCL. While they have 20 on occasion cited Intellectual Property Rights as 21 a reason for refusal, we are becoming increasingly 22 suspicious that the real reason is that the right level 23 of documentation simply has not been developed." 24 Was that a conclusion that you agreed with? 25 A. I would, yes. 131 1 Q. Were you one of the people that were interviewed that 2 said that to Project Mentors? 3 A. I can't remember whether I was interviewed for this 4 specific report. As I say, I'm confused between the 5 fact that there were two different Project Mentors 6 activities. So I couldn't be certain I was interviewed 7 for this one. 8 Q. The next paragraph: 9 "Of particular concern is the EPOSS system. We are 10 informed that at a relatively early stage Pathway wanted 11 the Authorities, principally POCL, to be involved with 12 the design of this element. The plan was to use the 13 Rapid Application Development ('RAD') methodology to 14 design this system. This approach was started, but 15 discontinued after some months, when the Pathway staff 16 member involved left the project. The suggestion to use 17 RAD leads us to believe that more traditional methods 18 have not been used, and since the RAD experiment was 19 abandoned, we have doubts whether any proper 20 requirements analysis has been performed." 21 Did you know about what's described in that 22 paragraph? 23 A. Yes, in that we knew that they had -- they wanted to 24 take this RAD approach and we knew the Pathway staff 25 member leaving. That was in my report that we saw a few 132 1 minutes ago. 2 Q. Lastly, "Impacts on the Programme in the Future". They 3 say: 4 "Our experience of systems where requirements have 5 not been analysed satisfactorily is that the system 6 fails to meet the users' needs. An effective acceptance 7 test will identify many such failings necessitating 8 considerable rework. The result is a significant 9 extension of time and cost required to complete the 10 system and roll it out. The alternative is to allow 11 unacceptable processing in the operational environment, 12 with unpredictable and potentially damaging results. 13 "In our opinion the failure to satisfactorily 14 analyse the requirements for the Benefits Agency makes 15 it unlikely that the users' needs will be met by the 16 current and Pathway system." 17 Was that a conclusion with which you agreed at the 18 time? 19 A. Yes. 20 Q. Can we look at your notes please, WITN05970121. You say 21 at the top: 22 "We have no evidence to disagree with the underlying 23 message in this report." 24 Who is the "we" in that sentence? 25 A. I believe I discussed this with other members of the 133 1 assurance team, that may have included John Meagher, it 2 may have included a couple of the Benefits Agency people 3 who were involved at the time as well. 4 Q. Can we look, please, at a couple of the points that you 5 made. The second bullet point from the bottom: 6 "Our view over the past 2+ years of experiencing 7 Pathway development has been: 8 "Horizon denied visibility, especially of 9 application design (oddly, we have managed to get some 10 visibility of the 'risk areas' such as the Riposte 11 middleware). 12 "no opportunity for Horizon to perform independent 13 [verification and valuation] 14 "no evidence that Pathway have performed 15 a [verification and validation]." 16 Those two points there, do you, by putting the 17 symbol to the left of them, mean that the result of what 18 you describe is that? 19 A. Yes. 20 Q. What do you mean by this bullet point as a whole? 21 A. Basically the same message as discussed earlier that we, 22 Horizon, as in the BA/POCL team had been denied 23 visibility of the application design. When I say we had 24 some visibility of the risk areas, what I mean by that 25 is the areas where we had raised formal risks to the 134 1 service provider risk creditor at the start, such as 2 Riposte. 3 In those cases, we did get more information but as 4 far as the application design, in particular for EPOSS, 5 we had been denied visibility. Therefore, we had no 6 opportunity for the programme staff to independently go 7 through the design and validate it against the 8 requirements. 9 What was maybe of more concern there is that, if 10 they didn't have design documentation at that point, 11 then how could they have done it? Bearing in mind it 12 wasn't our job to do it, if you like; it was their job 13 to do it and us to try and get content that they had 14 done it. That's not, in any way attempting to duck it 15 but they were the people in the PFI contract who were 16 doing this. If they didn't have the design 17 documentation, then, how could they assure the integrity 18 and correct operation of components such as EPOSS. 19 Q. Can we go over the page, please, and look at the first 20 full bullet point: 21 "Pathway adopting a 'end of pipe' approach to 22 quality and performance -- a 'fix on fail' in testing 23 approach, rather than putting in the effort to get it 24 right first time." 25 "End of pipe" approach? 135 1 A. By that I meant -- that's a term I've used elsewhere in 2 IT. It means rather than trying to do it right all the 3 way through the process, through a formal process of 4 requirements, and then design, and then having all the 5 quality processes in place; it was more getting to the 6 end and then -- 7 Q. See what flows out of the pipe? 8 A. Yes, and the "fix on fail" is something that we actually 9 commented on during the acceptance process as well, 10 where our concern was that there was a process of "Oh, 11 here's another bug, we will go and fix it", rather than 12 going back and looking at the fundamental problems that 13 led to that point. 14 Q. I understand an "end of pipe" approach to be an approach 15 to the treatment of waste that concentrates on treating 16 or filtering the effluent, I will call it, that flows 17 from a pipe, as opposed to making changes to the 18 processes that give rise to the effluent; is that a fair 19 way of describing it? 20 A. It is not one I have heard of before but I can see the 21 analogy, yes. 22 Q. Does that fit with what you were describing here as 23 Pathway's "end of pipe" approach? 24 A. It fits from the point of view that what we were seeing 25 from what they had exposed to us, they weren't -- the 136 1 effort wasn't being successfully deployed to get quality 2 performance, and whatever, right during the development 3 process. It was a matter, at the end of the pipe, when 4 the product came out of the end of the pipe, of then if 5 there were problems, then fixing it -- fixing the 6 individual problems. 7 Q. I understand, thank you very much. 8 Then can we look at the last bullet point on the 9 page, please: 10 "Pathway [you say] have generally shown an inability 11 or unwillingness to understand or recognise the complete 12 requirements set -- [see] the failure to follow the 13 (contracted) security standards, some whole requirements 14 missed (eg timeout back at R1c) etc. Have tended to 15 think they can develop a system without being bothered 16 by the detailed requirements or their meaning. Failure 17 to use the 'clarification' [methods]." 18 What did you mean by that last bullet point? 19 A. We had in place processes so that Pathway could bounce 20 questions back to the PDA or back to Horizon, in 21 particular on the application side, we had, at one 22 point, teams actually working in Feltham with Pathway, 23 team not of technical people but of business people, so 24 that people could be asked and they could have real 25 access to the business experts. 137 1 And certainly, at certain points in the process, 2 there didn't seem to be the level of clarifications or 3 questions coming. They wanted to get on and, in the 4 aforementioned black box, get on with developing it. 5 Q. In this document and some of the ones we have looked at 6 before, there has been mention of the EPOSS and you told 7 us a moment ago how critical the EPOSS was to this 8 system. Did you have any knowledge of the EPOSS task 9 force created or appointed by Terry Austin? 10 A. At the time, no, I don't believe I did. That is not to 11 say that there may not have been people working in 12 Feltham who did. Subsequently to leaving Post Office 13 and moving on, I have spoken to people who were involved 14 with Pathway at the time who -- on an informal, over 15 a pint basis -- who did indicate there was a time when 16 Pathway did pile almost as many people as they could get 17 into EPOSS to try and fix issues and the feedback I got 18 on that it wasn't that successful, they seem to be 19 introducing as many issues as they were fixing because 20 of the state of the code and because of the absence of 21 design. 22 So you can pile -- you can put people who may be 23 good but don't understand an area of code, they will 24 maybe fix one bug and introduce another one. I didn't 25 know it as a task force, less complimentary terms have 138 1 been used to describe it. I knew they had -- 2 subsequently, I had known they had piled people onto it. 3 Q. I think it follows that you weren't told of the work of 4 or the conclusions of the task force at the time that it 5 was undertaking its work? 6 A. Correct. 7 Q. Were you aware of any recommendations in 1998, 1999 or, 8 indeed, 2000 circulating within ICL that a rewrite or 9 redesign of EPOSS should be undertaken? 10 A. No. 11 Q. I think, as you have told us already, you have recently 12 been shown by the Inquiry a report on the EPOSS task 13 force. I wonder whether we could look at that, please. 14 It's FUJ00080690. 15 Is this the document that you referred to before and 16 after lunch -- 17 A. Yes. 18 Q. -- as having seen recently -- yesterday in fact. Does 19 it follow that you were not informed of the existence or 20 the contents of this 20-page report back in 1998, 1999 21 or 2000? 22 A. Correct. 23 Q. Ignore the date on the top right-hand side, 14/05/01, we 24 think that is an artifact of unknown origin. You will 25 see the abstract describes the report as describing the 139 1 activities of the EPOSS PinICL task force "which was in 2 place between 19 August and 18 September 1998 to reduce 3 to manageable levels the EPOSS PinICLs outstanding at 4 that time". 5 In terms of the individuals that are mentioned on 6 the distribution list there, can you help us as to your 7 understanding of the roles they were performing at, say, 8 September 1998, and Mr Austin? 9 A. Terry Austin was, I think, development director or 10 development manager. He was, I believe, responsible 11 overall for the software development. "M Bennett" would 12 have been Martyn Bennett, who was the director of risk 13 and security. Jan Holmes -- 14 Q. Just before you move on, D McDonnell, did you know -- we 15 know him to be David McDonnell. 16 A. No, I didn't come across him. 17 Q. Library, obviously, is not a person but either 18 a physical or an electronic repository. "Author: Jan 19 Holmes"? 20 A. I don't know if I ever met Jan but I believe he was the 21 audit manager or some quality and audit role, I think 22 probably working to Martyn Bennett. 23 Q. Thank you. Before we look into a couple of parts of the 24 report, what was your overall reaction when this report 25 was revealed to you? 140 1 A. Kind of annoyance and frustration because, if we had 2 understood everything that was in here, both from the 3 background and from what they did and what didn't 4 happen, I think it would have changed my attitude to 5 what we were doing. 6 EPOSS wasn't mine as such but if I had seen this, 7 I would imagine it would have been a subject of major 8 discussion in the PDA and the Horizon programme and 9 I think it would have had a major effect to what 10 happened one year later. 11 Q. What major effect may it have had a year later? 12 A. I think it may have made it far harder for acceptance to 13 have gone through. I think we would have -- I think if 14 we had known this, there would have been more pressure 15 on -- potentially to either do something to understand 16 better the state of the products at the point that it 17 moved from PFI to non-PFI and the BA left. I do think 18 that was maybe a missed opportunity. But I think if we 19 had known the state here, then, it would have been much 20 more evidence to a more cautious approach to what 21 happened in 1999. 22 Q. Can I take you to a couple of passages just before the 23 break, please. Look at page 6, please. You have read 24 this recently if there are any other parts you want to 25 draw to the Chair's attention, then do tell us, but if 141 1 we scroll down to "EPOSS Documentation (Section 7.1)": 2 "The document suite supporting the EPOSS product 3 consists of there main elements ..." 4 They are listed: 5 "All of these were developed by reverse engineering 6 the EPOSS product code at that time. What's the effect 7 or meaning of that passage in the report to you? 8 A. That tells me that design documents, high level and low 9 level design documents, could not have existed, 10 otherwise you wouldn't go and reverse engineer documents 11 after you had written the code. So that tells me that 12 the EPOSS product code had been written before there was 13 design documentation. 14 Q. So the suspicion that you had about withholding material 15 was correct? 16 A. This backs up the fact that it was withheld in this area 17 because it either didn't exist or didn't exist in a good 18 enough form to give us and if they said "Oh, we haven't 19 got any -- we have got to go up and re-engineer it", it 20 would have raised alarm bells. 21 Q. Over the page to page 7, please. Under "EPOSS Code 22 (Section 7.2)": 23 "It is clear that senior members of the Task Force 24 are extremely concerned about the quality of code in the 25 EPOSS product." 142 1 Just stopping there. Senior members of the task 2 force, do you understand that to be referring to ICL 3 itself? 4 A. Yes, I presume those were the most senior members of the 5 group of half a dozen, or whatever, that they brought in 6 for this task force. 7 Q. "Earlier this year the EPOSS code was re-engineered by 8 Escher and the expectation is that the work carried out 9 in Boston was to a high standard and of good quality. 10 Since then many hundreds of PinICL fixes have been 11 applied to the code and the fear is that code decay 12 will, assuming it hasn't already, cause the product to 13 become unstable. This presents a situation where there 14 is no guarantee that a PinICL fix or additional 15 functionality can be made without adversely affecting 16 another part of the system." 17 Was that something that you knew about at the time? 18 A. We knew that the EPOSS code had had a trip to Boston, we 19 were told, for completion. The rest of that, about the 20 state of the code and fears around code decay or 21 whatever, no. Again, if we had known the EPOSS product 22 was in that state, I think the project would have taken 23 a totally different path. 24 Q. Page 15, please. Scroll down, please. 7.1.1, under 25 "Documentation Suite". 143 1 "The EPOSS product was originally developed using 2 RAD techniques as part of the Joint Working Agreement in 3 force during Release 1. This approach carries a number 4 of attendant risks, not least of which is the lack of 5 formal specification. During 1997 the product was sent 6 to Escher for significant rework as the solution arrived 7 at via RAD was deemed not to provide sufficient 8 integrity." 9 Is that the same point? 10 A. That is the point that if you don't go through the 11 design process then, for something that is as complex 12 and critical as this, it is unlikely to have the 13 integrity it requires, which seems to be the case. 14 Q. Then, lastly, page 17, please. At the foot of the page 15 under "Existing Code": 16 "Although parts of the EPOSS code are well written, 17 significant sections are a combination of poor technical 18 design, bad programming and ill thought out bug fixes. 19 The negative impact of these factors will continue and 20 spread as long as the PinICL fixing culture continues. 21 This is partly due to the nature/size of the bug-fixing 22 task and partly due to the quality and professionalism 23 of certain individuals within the team." 24 If you had known about that at the time what, if 25 any, would the consequences have been? 144 1 A. I think if we had known about it at the time it would 2 have been suddenly escalated up to the highest level. 3 I think if we had known that this much existed, I am 4 sure the project would have potentially been canned, 5 come the point that the BA withdrew and the decision had 6 been taken of what was going to happen. I think this 7 would have tipped the balance. 8 MR BEER: Thank you. 9 Sir, is that an appropriate moment for the afternoon 10 break? 11 SIR WYN WILLIAMS: Yes, certainly. What's the likely 12 progress this afternoon, Mr Beer? 13 MR BEER: I'm likely to finish at about 4.30 pm or dribble 14 over into tomorrow morning and then some other core 15 participants will ask their questions tomorrow morning, 16 I suspect. 17 SIR WYN WILLIAMS: So there is no realistic prospect of 18 Mr Folkes finishing this afternoon? 19 MR BEER: I don't think so. 20 SIR WYN WILLIAMS: That being so, I think we should finish 21 at about 4.15 pm/4.20 pm, unless you really would like 22 10 minutes to finish, put it like that. 23 MR BEER: No, thank you, sir. 24 SIR WYN WILLIAMS: So 10 minutes now? 25 MR BEER: Yes, thank you very much, sir. 145 1 (3.16 pm) 2 (A short break) 3 (3.26 pm) 4 MR BEER: Good afternoon, sir, can you see and hear me? 5 SIR WYN WILLIAMS: Yes, I can. 6 MR BEER: Mr Folkes, can we turn to the position after or in 7 the run up to the new contract being signed between POCL 8 and ICL Pathway. 9 In your witness statement, you refer us to a paper 10 written by the late Keith Baines who was POCL's head of 11 commercial. 12 That summarises the contractual position in relation 13 to the new agreement that was due to be signed at the 14 end of July 1999. Can we look at that paper please. It 15 is WITN05970136. 16 You will see that it is a briefing paper on the new 17 Horizon contract. It is dated at the top in the box and 18 at the bottom right -- at the top, 27 July 1999, and at 19 the bottom, 27 July 2022. 20 Did you see this at the time back in July 1999? 21 A. Yes. This was a document that Keith produced and 22 circulated to the programme as it then was. I presume 23 the date on the bottom right-hand is just because that's 24 when I printed it off. 25 Q. When you printed it off. And it was just a coincidence 146 1 that it was 23 years later? 2 A. Yes. These things happen yes. 3 MR BEER: I'm sorry. I'm told, sir, that the transcript is 4 down. Thank you very much. (Pause) 5 We just have to pause for the moment because the 6 transcript has ceased to work. 7 (Pause). 8 Sir, I'm told that we need to break for five minutes 9 whilst a fix occurs. 10 SIR WYN WILLIAMS: Fine. I will go off screen but I will 11 still be in the room so just speak to me when it is 12 ready, all right? 13 MR BEER: Yes, thank you very much, sir. 14 (Pause for technical reasons) 15 MR BEER: Sir, I understand we are back up and running. 16 SIR WYN WILLIAMS: So I understand, Mr Beer. I was just 17 responding to an email, to the effect that I was ready 18 as well. 19 MR BEER: Thank you very much. Good to hear it. 20 Can we look please at the introduction of this 21 document at the foot of the page which describes its 22 purpose: 23 "This paper is prepared as a summary of the main 24 differences between the old and new contracts between 25 POCL and ICL Pathway, following the cancellation of the 147 1 benefit payment card, and the new commercial terms 2 agreed by the Post Office board on 24 May 1999. The 3 intended audience is POCL managers who need to 4 understand the main provisions of the contract either 5 because they have working contacts with Pathway or 6 because they have an interest in the services Pathway 7 will be delivering under the new contract." 8 You fell within that description? 9 A. Yes. 10 Q. Over the page please: 11 "On 24 May ... [POCL] and ICL Pathway signed 12 a letter agreement that made major changes to the 13 previous contracts between POCL and Pathway. At the 14 same time, DSS and ICL Pathway agreed to terminate their 15 contract for the benefit payment card services and DSS 16 withdrew from the tripartite Authorities Agreement. 17 "The letter agreement specified that POCL and 18 Pathway should produce a 'codified' contract that 19 incorporated the changes into a new contract by 16 July. 20 The codification process has now been completed, and the 21 new contract is due to be signed on 28 July." 22 That was the next day, after this document: 23 "There are significant changes to many aspects of 24 the contract, summarised in the following sections." 25 Would you, as a team leader or as a manager, have 148 1 used this document rather than the underlying 2 contracts -- contract itself to understand the relative 3 contractual obligations of each of the parties to it? 4 A. Certainly. But you would not have gone line-by-line 5 through the new contract. We would have used Keith's 6 summary. 7 Q. We can see in the rest of the document, I'm not going to 8 go through it now because it speaks for itself. If we 9 just scroll through it, differences in service scope, 10 difference in prices, it is modeled on a standard 11 turnkey build and operate contract, rather than a PFI 12 based usage charging structure, risk transfer: 13 "The new contract significantly reduces the risk 14 that Pathway was previously taking under the PFI. In 15 particular, risks in relation to achievability by POCL 16 of the roll-out timescales and in relation to future 17 business volumes are now largely with POCL." 18 Then, over the page please to "acceptance". About 19 five lines from the bottom of the first paragraph: 20 "The thresholds for medium severity Acceptance 21 faults have been increased from 10 faults in total, to 22 20 faults overall, or 10 in any one Acceptance 23 Specification. The threshold of any 1 high severity 24 Acceptance fault still applies." 25 So that amounts to a relaxation, is that right, the 149 1 increase in number of permissible medium severity 2 acceptance faults? 3 A. Yes, relaxation from the point of view of Pathway. 4 Q. Say again? 5 A. From the point of view of Pathway. 6 Q. Yes. Now, the product was designed and the system was 7 designed and developed under a PFI contract when, as you 8 have explained to us, you had limited visibility of what 9 was inside the black box and that was explained by ICL 10 Pathway on the basis that that was the nature of the PFI 11 contract and they were taking the risk under the PFI 12 contract. That's fair? 13 A. Yes. 14 Q. Now the relationship had fundamentally changed hadn't 15 it? 16 A. It had but we of course got to this stage with the 17 products, in theory, we were told, completed and almost 18 ready to go into acceptance. So it ceased to be a PFI 19 but it was starting from a very strange place to 20 a conventional contract. 21 Q. But the paragraph we looked at about the transfer of 22 risk to Pathway that existed under the PFI, that had 23 gone? 24 A. Yes. 25 Q. And now achievability by POCL of the rollout timescales 150 1 and future business volumes. Those risks now rested 2 solely with POCL? 3 A. Yes. 4 Q. To your knowledge, were the concerns, that you, in the 5 documents that we have seen, had been persistently 6 raising about technical aspects of Pathway's system, 7 brought into account in the renegotiation of the 8 contract? 9 A. I don't think they were, no. 10 Q. Do you know why? 11 A. I don't but I could hazard a guess but I don't know -- 12 I have no evidence on which to base it. 13 Q. If we look at your witness statement please at 14 WITN05970100 at page 42, you say at paragraph 124: 15 "The new Contract was based on the same set of 16 Requirements and sadly did not improve our ability to 17 perform Assurance on the solution. By the time that the 18 new contract was finally signed in mid-1999, the system 19 was already claimed to be fully developed ..." 20 That is the point you just made a moment ago? 21 A. Yes. 22 Q. "... and in many ways therefore 'the damage was done' 23 regarding assurance (and I fear Quality). The Contract 24 was signed as Acceptance activities were taking place 25 and the contract had no facility for us to go back and 151 1 revisit the problems of the previous few years, nor, as 2 far as I can remember, did it give us any further access 3 to detailed documentation." 4 To your knowledge, was any attempt made to stop this 5 from happening: either a failure to improve your ability 6 to perform assurance on solution, revisiting the 7 problems of the previous few years, or giving you 8 further access to detailed documentation? 9 A. I have no memory of any visibility of that. The 10 contract negotiation was going on -- off at a totally 11 different level and I don't think we were consulted as 12 to what, if anything, we wanted to get in there. 13 I think it was being done at some speed and went through 14 some external pressures as we have seen and I don't 15 think -- it was never seen to be on the table, you know. 16 Nobody coming round and saying "Now, what else should we 17 put in here". 18 Q. It might be suggested that there was an unholy rush, 19 amongst some, to get Horizon done. What would you say 20 to that? 21 A. There was a lot of pressure to get Horizon done. 22 Obviously the programme had been going for five years. 23 There was a big expectation out in the network, the 24 network of offices. There were lots of -- no, he's not 25 here -- a glacier being held back from coming down the 152 1 valley. 2 Everything was poised from the point of view of the 3 big rollout of 20,000 offices. So there was a lot of 4 pressure on POCL management for it to be done. I think, 5 from what we see around acceptance, there was an attempt 6 during the acceptance period to put the brakes on when 7 the number of Acceptance Incidents came out and some of 8 the issues came out. I think at the point view of -- 9 point of time of signing the contract I think there was 10 probably a lot of pressure to get the contract signed 11 for political and commercial reasons and probably no 12 desire to, sort of, reopen the box by changing the rules 13 as it were, however beneficial that might have been. 14 Q. In paragraph 126 of your statement, if we just scroll 15 down please, you say: 16 "In hindsight, POCL missed an opportunity to force 17 an in-depth review of the emerging system, including 18 examining all aspects of how the system had been 19 created, to mitigate the gap caused by the earlier PFI 20 approach, and to require Pathway to open up to POCL." 21 You say that in hindsight POCL missed 22 an opportunity. Was hindsight, in fact, necessary in 23 order to reach that conclusion or was it obvious at the 24 time? 25 A. I think -- I'm just trying to give you the most honest 153 1 answer I can to that. I think after three or four years 2 of -- at times it has felt like fighting with Pathway, 3 from the point of view of assurance etc -- that, to me, 4 it was obvious that we should have done something but 5 there was an incredible amount going on at that point, 6 from the point of view of the new contract and whatever. 7 I don't know whether we would have got anywhere if we 8 pushed harder. If [coughing] and myself had pushed 9 harder to get an in-depth review. I think it was, 10 a sort of, unstoppable super tanker at that point. 11 There was a view I think, that, okay, we have got 12 through all this period, the assurance process hasn't 13 been a great success for all the reasons we have spoken 14 about for several hours but now we are coming up to 15 acceptance and people were going to rely upon the 16 acceptance process and the acceptance testing and 17 Acceptance Incidents to try and ensure the quality of 18 the system. 19 I don't know if that answered the question or not. 20 Q. Well, you had been identifying and reporting in writing, 21 serious concerns and we have seen some of them today. 22 We don't have a record of what you were saying orally or 23 via email. Sometimes they are even stronger in 24 expression. But we have seen the formal documents that 25 you created. Was this issue overlooked, ie we have not 154 1 been able to undertake an in depth review of the 2 emerging system, including examining all aspects of how 3 the system had been created or was it known about but 4 there was a rush to get the system rolled out? 5 A. I don't think anybody could claim that it wasn't known 6 about. You know, from the number of documents, given 7 the amount that myself and others on the team had said, 8 and they probably got rather bored with us saying it, 9 but we had been very clear on the message, so I do not 10 think it was not known about. 11 I can't really explain why people didn't go out and 12 say "Okay, what else should we be doing at this stage". 13 The contracts team -- renegotiating the contract and 14 their brief, I guess, was to take the contract and 15 codify it down to the single contract. You say, was 16 there an almighty rush? From what I have seen 17 understood, there was considerable pressure being put on 18 POCL to be able to move this forward, which probably 19 would have not encouraged, "Yes, reopen the contract and 20 see what other quality and access requirements we 21 wanted". 22 Q. Thank you. Can I turn to a new topic. Audit, the 23 provision of information and data by ICL Pathway and 24 prosecutions. Can we look please at your witness 25 statement WITN05970100, at page 67, please. This is 155 1 within a section of your witness statement when you are 2 looking back at fitness for purpose at rollout and at 3 paragraph 207, in the third line you say: 4 "I think a key question for the Inquiry to ask is 5 what gave POCL such confidence in Horizon to start using 6 it for prosecutions of Subpostmasters, especially after 7 the rather chequered history of the system from 8 1996-2000 and in particular the experiences of 1999." 9 Can we look please, in addressing that question, at 10 Fujitsu, FUJ00000071. This is the codified agreement of 11 July 1999. Can we turn to page 97, please. Can we 12 look, please, at 4.1.8 and 4.1.9. 4.1.8: 13 "The Contractor shall ensure that all relevant 14 information produced by the POCL Service Infrastructure 15 at the request of POCL shall be evidentially admissible 16 and capable of certification in accordance with the 17 Police and Criminal Evidence ACT ... 1984. The Police 18 and Criminal Evidence (Northern Ireland) Order 1989 and 19 equivalent legislation covering Scotland. (R829 20 para 1)." 21 What do you understand the reference to "R829 para 22 1" is? 23 A. That was a requirement in the requirements catalogue, 24 829.1. 25 Q. It is part of the specification in the requirements yes? 156 1 A. Yes. 2 Q. A cross-reference, if you like, to that? 3 A. Yes. 4 Q. 4.1.9: 5 "At the direction of POCL, audit trail and other 6 information necessary to support live investigations and 7 prosecutions shall be retained for the duration of the 8 investigation and prosecution irrespective of normal 9 retention period of that information", and then 10 a cross-reference to R829.2? 11 A. Yes. 12 Q. Now, I think it is right that you were probably aware of 13 these two provisions at the time, ie back in 1989 -- 14 sorry, 1999? 15 A. Yes. In the same way I was aware of most of the 16 provisions in the requirements. 17 Q. The reason I say that is we are going to come in 18 a moment to a report that you wrote in the late summer, 19 early autumn of 1999, about the subject of compliance 20 with these two provisions. 21 Before we get there, can we look please at 22 POL00029165. Thank you. 23 Can you see this is a document described as the 24 Horizon system audit manual? 25 A. Yes. 157 1 Q. Can you please summarise at a high level what it is, 2 please? 3 A. A document created by Pathway and it was shared with -- 4 this version -- with Post Office Internal Audit, who is 5 POIA. It had previously been -- an earlier version that 6 I had referred to the Inquiry to was shared with other 7 members of POCL and BA Audit and it basically describes 8 in some detail the methods of access to audit material 9 by various different groups of auditors, including 10 Pathway internal auditors, POCL auditors, BA auditors 11 and some other categories of auditors and, in one 12 particular part of it, it does attempt to address the 13 requirements of 829, which is the one that you just 14 referred to -- 15 Q. Thank you, we will come to that in a moment. But 16 I think this is the first time we have looked at this, 17 so I just want to introduce it slowly. You see the 18 reference in the top right, "IA/MAN/004". Does that 19 refer to "Internal Audit/Manual"? 20 A. I believe that was Pathway's naming convention, yes. 21 Q. The version we are showing you is 1.3, dated 22 17 January 2000, so shortly before you left in 23 February 2000. That's why we have picked this one to 24 show you. 25 A. It was a much earlier version, 0.3, which is the only 158 1 one I believe I had ever seen and that was one that 2 still had reference to BA on it because it was pre the 3 split. 4 Q. Just looking at the abstract: 5 "[The] manual describes the Horizon Operational, 6 Operational Support and Commercial systems and data 7 flows in sufficient detail to enable members of the 8 Horizon Audit Community to understand then for audit 9 purposes. 10 "It also addresses the appropriate Criteria of 11 Requirements", and then some numbers given, including 12 829. That's the cross-reference we looked at earlier. 13 "... insofar as it provides information relating to 14 the composition of and access to the 'audit trail' as 15 defined in those Requirements and its admissibility for 16 PACE certification." 17 So the abstract -- the second part of it is probably 18 the most important part for us -- is saying that this 19 document addresses requirement 829? 20 A. Yes. 21 Q. If we scroll down the distribution list. Can you just 22 tell us where within each organisation people came from, 23 on the distribution list? 24 A. Martyn Bennett was the Pathway director of risk and 25 security. Chris Paynter was from Post Office Internal 159 1 Audit. 2 Q. So POIA? 3 A. I think it is Post Office Internal Audit. 4 Q. What did Post Office Internal Audit do? 5 A. They were the internal audit function of the Post 6 Office. I didn't have any detailed contact with them 7 but they would have been considered to be the sort of 8 expert domain. The programme sat in a role of almost 9 a dating agency, if you like, bringing in the relevant 10 expert domains from around the organisation, together 11 with the expert domains in Pathway. So, in this case, 12 Chris Paynter, as POIA, was -- I don't know where he sat 13 in internal audit but he would have been their nominated 14 contact and the nominated representative of the audit 15 community in Post Office who would have needed access to 16 the system. 17 Q. What relationship did Post Office Internal Audit have 18 within the Post Office or Royal Mail, indeed, for the 19 conduct of investigations and prosecutions? 20 A. I can't answer that one. 21 Q. Were POIA responsible in any way for investigation or 22 the conduct of criminal investigations? 23 A. I can't answer that one, I'm afraid. 24 Q. Then library, do you know whose library that is? 25 A. In all these documents, "library" means the Pathway 160 1 library. 2 Q. So a joint library? 3 A. No, I believe that is the library run by Pathway. 4 Q. ICL Pathway. 5 A. I don't know whether this were agents to share certain 6 documents but, certainly, there are plenty of documents 7 that say "Library" that weren't shared. 8 The other name on there, Paul Redwood -- 9 Q. Yes. 10 A. -- was a member of the POCL Horizon team. 11 Q. What position did he hold? 12 A. The name doesn't really ring a bell. I presume it may 13 have been within -- I couldn't really make a good guess. 14 Q. So it is an ICL Pathway document. 15 A. Yes. 16 Q. What do you know about POCL's contribution to it or 17 review of it? 18 A. There was an earlier version of it and there were two 19 POCL people on it, two other names. One, I think, was 20 a Hilary Stewart and the other was a Jason Carter(?). 21 I don't personally know or didn't know either of them 22 but I presume they represented the audit community and 23 I believe Hilary Stewart contributed comments in this 24 document in that there is reference to comments from 25 "HS" further down in the change history. 161 1 Q. Yes, if you go over to page 3, please. At the foot of 2 the page, I think we can see, under "Changes v1.2 to 3 1.3", for some reason Hilary Stewart and a lady called 4 Ruth Stinchcombe were removed from the distribution. In 5 any event, this is a 68-page document, and I'm only 6 going to draw your attention to some of it. 7 Can we turn, please, to page 53. We should just 8 look at the previous page -- I'm sorry -- to get the 9 heading, "Retrieving and Extracting Audit Data", 10.3 10 and then scroll back to where we were. 11 "Overview", I'm going to ask you to translate this 12 in a moment: 13 "This is where audit data is retrieved from the DLT, 14 based on Request(s) For Information made by Post Office 15 Internal Audit, and presented for further extraction or 16 placed on CD-ROM or other suitable media for dispatch to 17 the [Request For Information] originator. 18 "The following paragraphs are ordered to reflect the 19 actual processing of a Request For Information (RFI) by 20 ICL Pathway Internal Audit." 21 Can you just translate what that's saying, please? 22 A. DLT stands for digital something-or-other tape. It is 23 a storage medium. So my understanding is that audit 24 data would -- if you go up to the previous diagram, if 25 we can just go up one page, I believe we should see 162 1 there are tapes on the left-hand side. 2 So data gets written to tapes. I believe, Pathway 3 had an infrastructure that pulled data from a number of 4 points in the system, including of interest to us here, 5 what's called the TMS journal, which is the transaction 6 management service journal, the data coming up from 7 offices. That data was stored on DLT, from what 8 I remember, probably on a daily basis, and what the 9 Pathway internal audit can do is retrieve that data as 10 per request and copy the data onto CD-ROM, compact disk, 11 to give to an authorised Post Office Internal Audit 12 person. 13 Q. Back to the next page, please. 10.3.2: 14 "POIA [Post Office Internal Audit] will request 15 audit data via Request For Information form. This will 16 contain a description in business terms of the times 17 outlets, events, items and activities that the Auditors 18 are interested in. This request has to be interpreted 19 by Pathway Internal Audit and mapped onto the Audit 20 Points and Files described earlier in this manual." 21 I'm not going to take you to those. 22 Can we go over the page, please, to 10.4.2, which 23 may be an important paragraph, "Investigation Support": 24 "The term 'investigation' is used in its broadest 25 sense and does not limit itself to fraud. Any RFI is 163 1 likely to be associated with a specific business event, 2 eg an encashment, a bill payment, an outlet, 3 a beneficiary. It is anticipated that the majority of 4 this type will be based on the TMS journal or will use 5 it as a start point. See section (11.2) [which we can 6 look at] for details of how to raise a RFI." 7 Can you translate what that's saying for us, please? 8 A. The statement is basically just saying that 9 investigation doesn't necessarily mean fraud 10 investigation but anything that POCL need to 11 investigate. Typically, that would be -- you would be 12 able to start by -- it may be a specific event, such as 13 a specific transaction or in a specific outlet that 14 would trigger it, and it's saying the majority of this 15 type of retrieval would be based on the TMS journal. 16 The TMS journal is, I believe, the term they use for 17 the message store centrally which is, by nature of 18 Riposte, the image of the message store in the office. 19 So that's, if you like, the raw data that EPOSS would 20 have written in the office, once it is replicated up to 21 the centre, it is then, I believe, archived onto tape 22 from there and it is bringing back that data from tape 23 to give the journal that would then be available for the 24 investigation. 25 Q. Thank you. Over the page, please. "Obtaining Access to 164 1 Operational Audit Data", this is under the heading 2 "Requirement 699". As far as I can see, there isn't 3 a specific heading dealing with requirement 829, which 4 is the one that we were interested in. 5 But under requirement 699, if we scroll down, 6 please, to the bottom of the page, there is something 7 that may be of interest, at the last part on the page: 8 "Access to POCL audit trails, particularly the TMS 9 Journal, is seen as a strict POCL preserve. If any 10 third parties require access to it, for evidential 11 purposes or fraud investigation, then the access will be 12 via Post Office Internal Audit." 13 A. If I can just explain what's behind that paragraph. 14 Q. Yes, please. 15 A. Apart from, if you like, the obvious, which is that 16 PO Internal Audit is the conduit into Pathway, when it 17 was still a joint contract, there were many joyous 18 debates between BA and POCL regarding access to 19 information and, if you had a potential -- mostly around 20 benefit encashment fraud. If there had been benefit 21 encashment fraud down on a counter terminal, the 22 question then came up is that data BA's data or POCL's 23 data, because it is sitting within the POCL data but it 24 may relate to the a BA transaction. 25 I think all this was actually saying is that, if it 165 1 was data down on the POCL counter terminal, then access 2 to that data, even if it referred to a BA transaction, 3 would need to go via POCL. 4 Q. Looking at the document as a whole -- and I think you 5 had a chance to read it and I think I have drawn your 6 attention to those parts of it that could or could 7 possibly be a reference to carrying requirement 829 into 8 effect -- what's your view as to whether this document 9 adequately addresses, from an operational and practical 10 level, the translation of 829 into reality? 11 A. Obviously, the document doesn't make any reference to 12 any form of certification or any form of -- doesn't 13 cover anything apart from direct access to the data. 14 Q. Sorry, just stopping there. You are making the point, 15 I think, that the document is focused on obtaining 16 access to data, rather than the provision of a statement 17 or a report that certifies accuracy or integrity; is 18 that right? 19 A. Yes, and I think this is very much looking at it from 20 an operational/technical point of view, so it goes into 21 great detail on, if you like, the integrity of the audit 22 trail, how it is pulled off, et cetera, but not around 23 anything other than, say, the provision of statements or 24 whatever, as may be required by PACE. 25 I don't know where that would have been -- other 166 1 part would have been covered. 2 Q. Is it right that you took some steps to try and get that 3 other part, ie compliance with PACE, covered? 4 A. So, I believe in the original contract, there were the 5 same -- those two requirements but, as applied in the BA 6 contract, they were slightly different. 7 Q. We are going to come to that in a moment. Just, 8 generally, did you make some effort to try and ensure 9 that requirement 829 was translated into practical 10 effect so that systems and processes were created at 11 a practical level to ensure that it could occur? 12 A. I didn't personally, but, without dodging the question, 13 requirement 829 and audit wasn't part of my remit. 14 Q. Did you write a paper suggesting it be done? 15 A. We did -- there was one piece of work I was asked to 16 look at in the middle of -- September 1999, where I was 17 passed, I think by possibly Dave Miller or Bruce 18 McNiven, a document that Post Office Internal Audit had 19 written that made various comments about audit and 20 investigation access, and I did respond back to either 21 David Miller or Bruce McNiven around that and it did 22 cover the fact that, in particular, there seemed to be 23 a gap regarding the provision of a PACE -- I don't want 24 to use the term loosely -- but witness statement and 25 that that seemed to be something that had been dropped 167 1 out in the contract and would need to be taken forward. 2 Q. Thank you. 3 A. So I was trying to flag I'm not an expert in this area 4 but this is something that should need to be resolved. 5 MR BEER: Sir, as we are on this topic, would you mind if 6 I just spent 10 minutes finishing it off? 7 SIR WYN WILLIAMS: No, you carry on, Mr Beer. 8 MR BEER: Can we look, please, at that paper, WITN05970134. 9 If we can just blow it up a little bit, thank you. 10 Just scrolling down to start with -- if you can 11 scroll down, please? 12 A. Can I say, the bits that are in boxes are my comments. 13 Q. That is what I was about to ask. 14 A. The bits that are in boxes are my comments. I was 15 passed this paper by either, I think, Dave Miller or 16 Bruce McNiven and asked to comment on it and I -- my 17 comments are here in the boxes. 18 Q. So bits in the boxes are your writing, bits outside the 19 boxes are part of the pre-existing content of the 20 document? 21 A. Yes, which was, I believe, written by somebody in Post 22 Office Internal Audit but I don't have their name. 23 Q. Can we go up to the top of the page, please. Under the 24 "Introduction": 25 "A review of the Horizon Cash Account System was 168 1 undertaken following a request from the Horizon 2 Programme Director. The objectives of the audit were 3 reflected in the terms of reference which were agreed 4 with him on 15 July 1999." 5 At about that time, July 1999, who was the Horizon 6 programme director within POCL? 7 A. I presume that was still Dave Miller but it would either 8 have been Dave Miller or Bruce McNiven. 9 Q. "It was agreed that this review would be undertaken in 10 two stages and this report reflects the findings from 11 the second stage. 12 "Management Summary 13 "POSIS Investigations at Outlets." 14 Does "POSIS" stand for Post Office Security and 15 Investigation Services? 16 A. I believe so, yes. 17 Q. "We are extremely concerned to be informed during the 18 review that POSIS currently do not have access to 19 archived data from the system. Data on the system is 20 compressed and archived after 35 days. It was 21 originally intended that access would be gained via the 22 Fraud Risk Management Server, which formed part of the 23 benefits payment system and has now been withdrawn. 24 This means the business could be in a position where it 25 is unable to investigate potential frauds or prosecute 169 1 cases due to the unavailability of critical data." 2 Stopping there. You explain in your witness 3 statement the inclusion of the FRMS, the fraud risk 4 management server, in the original contract, the 5 tripartite arrangement, and its withdrawal when the new 6 contract was signed in July 1999? 7 A. Yes. 8 Q. Have you any comment to make on that paragraph there? 9 A. Well, the paragraph shows a degree of confusion amongst 10 the writer. The fraud risk management service was one 11 of the components that was contracted for by BA. It was 12 created specifically for the Benefit Encashment Service 13 or Benefit Payment Service and the idea was that, as far 14 as introducing the payment by plastic card to coming on 15 for 20 million benefit recipients, there would need to 16 be potentially control of encashment fraud. 17 And the FRMS was a service that I think was in the 18 requirements and that Pathway posed a solution for, to 19 have specific reports and mechanisms to flag up 20 potential fraud within the encashment process. 21 As examples of this, I went back and checked. One 22 of the examples was, for instance, if an office did more 23 than its expected number of foreign encashments -- 24 a foreign encashment being where somebody was away from 25 home -- you might get an office if it -- suddenly find 170 1 that its normal list of number of foreigns was 2 5 per cent, it was actually 80 per cent, it may indicate 3 fraud. 4 There was also something called casual agents, where 5 people could go in without a card and if an office got 6 a higher number on that, it might indicate fraud. 7 The FRMS was very much intended not for prosecution 8 support, not to get audit data, but as a means of 9 managing benefit encashment risk and identifying high 10 risk activities. 11 When BA withdrew, FRMS ceased to exist because it 12 was a BA service. It also wouldn't have been relevant. 13 I think there was only one report, which something about 14 out of hours transactions, but everything else was 15 benefit encashment specific and, therefore, it would 16 have had nothing to do. It was never intended that 17 POSIS or any of the Post Office audit or investigating 18 people would go in via FRMS to get this kind of data. 19 Q. Over the page, please. Do you explain that -- I'm not 20 going to go through it -- in paragraphs 4, 5 and 6, 21 where you begin "There appears to be some confusion"? 22 A. I do, yes. And just to say, the only thing I would say 23 is on para 5, what I'm saying is there were quite 24 reasonable arrangements between POCL Security and 25 Investigations and the BA equivalent, who were called 171 1 Organised Fraud, to share data because, if there was 2 fraud regarding benefit encashment, it may be 3 appropriate to share it. But it was never intended that 4 POCL would have to go to BA Organised Fraud to get 5 information on an office having a problem with its cash 6 account or doing a bill payment transaction. 7 Q. Thank you. If we just move down to paragraph 7, please. 8 You say: 9 "A documented process has been established between 10 POCL and ICL Pathway for access to historical audit 11 information from Pathway's central systems, under the 12 auspices of the requirements on Audit (Requirements 699 13 and 829), using an [RFI] procedure. This process 14 involves the exchange of the RFI and data between 15 nominated individuals in the audit domain, believed to 16 be (originally) Hilary Stewart and (now) Chris Paynter 17 of POCL National Audit, and Jan Holmes of ICL Pathway's 18 Audit team. All requests from POCL would therefore need 19 to be fed through the nominated POCL audit contact. 20 This process is documented within the Horizon System 21 Audit Manual (IA/MAN/004)." 22 That is the audit manual we just looked at? 23 A. Correct. 24 Q. Over the page, please. The original authors wrote that: 25 "Bob Martin also advised us that Security and 172 1 Investigation Executive (S&IE) had requested an expert 2 witness statement from Pathway to support a prosecution 3 and that this had been refused on the grounds that there 4 was no contractual requirement. John Cook advised us 5 that there is a contractual requirement for Pathway to 6 ensure that the system meets the requirement of the 7 Police and Criminal Evidence Act. There is a need for 8 Pathway to agree with the [Security and Investigation 9 Executive] and Internal Audit how this requirement will 10 be met, as well as the procedures for obtaining this 11 evidence when needed for prosecutions." 12 Then you, in your box, say: 13 "The requirement (R829.1) is actually", and you set 14 it out, yes? 15 A. Yes. 16 Q. "An outstanding [AI] Acceptance Incident (370 Low) -- 17 exists against the POCL requirement, on the assertion by 18 POCL that Pathway should produce a witness statement to 19 support prosection. This AI revolves around the 20 interpretation of 'ensure that all relevant information 21 is evidentially admissible' -- POCL's view is that to be 22 admissible it will need to be supported by witness 23 statements etc; Pathway have stated that they will 24 'provide PACE statements as necessary to support a fraud 25 prosecution', but that 'the work required to produce 173 1 draft witness statements' is not within the scope of the 2 requirement and will be done once POCL raise a Change 3 Request. 4 "This issue has been handled with Pathway by Bob 5 Martin and Paul Harvey of SIE. The Acceptance Incident 6 is still open, and its resolution would appear to be 7 primarily a commercial issue", ie to do with money? 8 A. Yes. 9 Q. Then you say this in square brackets and I think this is 10 what you are alluding to earlier: 11 "Note the Benefits Agency has similar requirements 12 (R741 and 780) covering their aspects of the service 13 mirroring the above section, however these had 14 an additional clause reading: 'The contractor shall 15 provide certification in accordance with ... PACE, and 16 PACE(NI) ... (and equivalent Scottish legislation) when 17 necessary for a proposed prosecution ...'" 18 Then this: 19 "'... to demonstrate that the Service Infrastructure 20 was operating within normal parameters at time of 21 an alleged offence'. This additional clarifying clause 22 was lost at the time of withdrawal of BA (at contract 23 codification), but in any effect it (strictly) only 24 referred to PAS and CMS, both BA services, even prior to 25 BA withdrawal. Attempts were made to get this cause 174 1 inserted into the codified agreement to apply to POCL 2 but without success. 3 "Clearly, a process does need to be agreed between 4 POCL, SIE and ICL Pathway for the commissioning of PACE 5 certification, statements and court appearances." 6 Firstly, how do you know that attempts were made to 7 get the previously worded clause inserted into the 8 codified agreement as it applied between POCL and ICL 9 Pathway but without success? 10 A. I do not know for certain but I imagine I went up and 11 spoke to John Cook -- John Cook, I believe we mentioned 12 earlier on -- John Cook worked within Keith Bates' team 13 on the contract. 14 Q. The last sentence: 15 "Clearly a process does need to be agreed ... for 16 the commissioning of PACE certification, statements and 17 court appearances." 18 To your knowledge what happened? 19 A. I don't know what happened to this, in that I was asked 20 to comment on this by, I think, Dave Miller and 21 I returned this back to Dave Miller, I believe. 22 Q. Lastly if we can look at page 5 please under conclusion. 23 The original author/s say: 24 "There is a need to ensure that the problems 25 relating to the audit trail for S&IE investigations and 175 1 demonstrating that the system meets the requirements of 2 [PACE] have been impact assessed as incidents and are 3 considered by the Acceptance and Release Authorisation 4 Boards if not satisfactorily resolved. In addition, it 5 will be necessary to consider whether the current level 6 of cash account errors will affect the accuracy of 7 settlement with clients, when considering the rate at 8 which the system should roll-out." 9 What is Mr Miller or whoever wrote this saying in 10 the last paragraph? 11 A. It wouldn't have been Mr Miller, it would have been the 12 auditor. 13 Q. Sorry. What is the auditor saying -- 14 A. The auditor in that paragraph is making the statement 15 that, with the level of errors that were being reported 16 at this point, which was September I think, I can't 17 remember the date at the top -- but that has been 18 reported in late summer of 1999, that it would -- that 19 that level of errors would affect the ability of the 20 POCL back end to manage settlement with clients. 21 Clients in this case means the companies for whom 22 POCL does work, such as, you know, BT or water companies 23 or British Gas or whatever. 24 Q. To your knowledge was anything done with the 25 recommendations made by this paper whether -- as 176 1 originally authored by the auditor or as amended or 2 commented on by you? 3 A. I didn't have any visibility of anything being done with 4 it, but I believe it would have been. I believe 5 something was put in place -- for better or worse, there 6 was a process put in place for the production of expert 7 witness statements, and I presume there was also -- 8 processes were better exercised as far as retrieval of 9 audit data, but I -- at that point, this is one of these 10 jobs that I was set the task by I believe Dave Miller: 11 comment on this paper. I comment on the paper and went 12 back. 13 This was about the time when a number of tasks were 14 being passed away from people on the programme, like 15 myself, who eventually would disappear, into Business 16 Service Management and I think a number of the 17 activities, as far as communication with Fujitsu or with 18 Pathway, as far as access etc, would have then moved to 19 being managed by Business Service Management. But 20 a likely place it might have ended up may have been in 21 BSM. 22 MR BEER: Thank you, Mr Folkes. It is 4.30 pm. I have 23 about 10 minutes more which I would propose to pick up 24 tomorrow morning. 25 SIR WYN WILLIAMS: Yes, fine. All right. We will start 177 1 again at 10.00 tomorrow morning then. 2 No doubt, Mr Folkes, you will be equally glad not to 3 speak about your evidence overnight. 4 A. I will go and find a restaurant where nobody knows me. 5 SIR WYN WILLIAMS: Fine. See you in the morning. 6 MR BEER: Thank you very much, sir. 7 (4.29 pm) 8 (The Inquiry adjourned until 10.00 am on Thursday, 9 3 November 2022) 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 178 1 2 I N D E X 3 MR STUART SWEETMAN (continued) .......................1 4 Further questioned by MR STEVENS .................1 5 Questioned by MS PATRICK ........................11 6 Questioned by MR WHITTAM ........................26 7 MR JEREMY PETER FOLKES (sworn) ......................31 8 Questioned by MR BEER ...........................31 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 179