Official hearing page

7 March 2023 – Elizabeth Evans-Jones and Amandeep Singh

Hide video Show video

(10.00 am)

Mr Stevens: Good morning, sir, can you see and hear me?

Sir Wyn Williams: Yes, I can, thank you.

Mr Stevens: Thank you, sir. This morning we will be hearing evidence from Elizabeth Evans-Jones, but earlier today a few further documents were provided for her to consider and I would ask that we could adjourn her evidence by 15 minutes to allow a bit more time for her to consider those documents.

Sir Wyn Williams: Yes, of course, Mr Stevens. So we will begin as soon after 10.15 as you think appropriate.

Mr Stevens: I’m very grateful. Thank you, sir.

Sir Wyn Williams: Right.

(10.01 am)

(A short break)

(10.16 am)

Mr Stevens: Sir, can you see and hear me?

Sir Wyn Williams: Yes, I can, thank you.

Mr Stevens: Thank you for the time. If I may call Ms Evans-Jones.

Elizabeth Evans-Jones


Questioned by Mr Stevens

Mr Stevens: Good morning. As you know, my name is Sam Stevens and I ask questions on behalf of the Inquiry. Please could you state your full name?

Elizabeth Evans-Jones: Elizabeth Jane Evans-Jones.

Mr Stevens: Firstly, thank you for giving evidence to the Inquiry, both in a written statement which we’ll turn to today and also for attending to give oral evidence, and considering the additional documents that we gave to you this morning. You, should have a written statement in front of you in the bundle of documents under tab A. Do you have that there?

Elizabeth Evans-Jones: I do, yes.

Mr Stevens: It runs to 21 pages.

Elizabeth Evans-Jones: It does indeed, yes.

Mr Stevens: On page 16, there should be a final paragraph 43 –

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: – and beneath that a statement of truth and your signature. Is that your signature?

Elizabeth Evans-Jones: I can confirm it is, yes.

Mr Stevens: Can you confirm that the contents of that statement are true to the best of your knowledge and belief?

Elizabeth Evans-Jones: Absolutely, yes.

Mr Stevens: Thank you. That stands as evidence in the Inquiry now and, for the transcript, the reference is WITN06680100. I am going to ask you some more questions about it but not cover everything that’s within it.

Firstly, by way of background, you graduated in 1999 –

Elizabeth Evans-Jones: I did.

Mr Stevens: – and you joined Fujitsu in October 2005 to work on the Post Office Account?

Elizabeth Evans-Jones: That’s correct, yes.

Mr Stevens: You were a Service Delivery Team manager?

Elizabeth Evans-Jones: I was, yes.

Mr Stevens: If you could just move slightly closer to the microphone, it’s just – thank you, I’m very grateful.

Elizabeth Evans-Jones: You’re welcome.

Mr Stevens: Prior to that role, could you summarise any qualifications that you had that were relevant to carrying out that job?

Elizabeth Evans-Jones: So I was ITIL certified Version 3 Expert, so that’s the IT Infrastructure Library, which delivers best practice in terms of how to manage and delivery services.

Mr Stevens: What work experience or professional experience had you had in delivering a role like that prior to joining Fujitsu?

Elizabeth Evans-Jones: So prior to joining Fujitsu, I worked in Threadneedle Asset Management where I ran the service management department for a period of time; I was also a change release and configuration manager; and previous to that, I worked for Yellow Pages also in ITIL service management functions.

Mr Stevens: When the role came up for you to join Fujitsu, do you recall how you were selected for it?

Elizabeth Evans-Jones: I was recommended by a colleague who used to work – I was put forward – who I used to work with at Threadneedle, he put me forward for the application to Fujitsu. I then went through two or three rounds of interviews with Fujitsu and then I was selected for the role.

Mr Stevens: You stayed in the role on the Post Office Account until December 2007?

Elizabeth Evans-Jones: That’s correct.

Mr Stevens: Then you moved to a different account but within Fujitsu?

Elizabeth Evans-Jones: That’s correct, yes.

Mr Stevens: When you were working for that different account, from December 2007 onwards, did you have any more working or dealing with the Post Office Account?

Elizabeth Evans-Jones: Not from a work perspective. Obviously I had colleagues that I interacted with but not from a work perspective.

Mr Stevens: So when you finished on the Post Office Account in December 2007, that’s your last dealings with the Horizon and the Post Office Account?

Elizabeth Evans-Jones: Correct.

Mr Stevens: You left Fujitsu in August 2010?

Elizabeth Evans-Jones: December – err, yes, August 2010, that’s correct, yes.

Mr Stevens: I want to look at support services generally first. You’re primarily going to talk about what was known as either the Horizon System Helpdesk or the Horizon Service Desk?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: I’m just going to refer to it as the Helpdesk for today –

Elizabeth Evans-Jones: Okay.

Mr Stevens: – and that was first line support.

Elizabeth Evans-Jones: Correct.

Mr Stevens: From an IT background, how would you describe the purpose of first line support?

Elizabeth Evans-Jones: From an ITIL perspective, it’s intended to be the single point of contact for clients to interact with an organisation, be that for software, hardware, or general queries. The desk should then log the incident, so that it’s captured from a volumetric perspective, attempt to troubleshoot and resolve at first point of contact, if not possible to resolve, then to refer that through to second or third line support, depending on the processes.

Mr Stevens: So one of the purposes is to try to resolve the issue at first line and then, if not possible, refer up to the second line or third line?

Elizabeth Evans-Jones: Yeah.

Mr Stevens: On the Post Office Account, second line support, we understand, was provided by the Systems Management Centre or SMC?

Elizabeth Evans-Jones: Depending on the nature of the incident, so second line for hardware faults would be potentially engineering services but for software, yeah, absolutely, through to the SSC, I believe the team was called.

Mr Stevens: So the SSC – we have heard about the SSC at third line. Do you recall the SSC, the System Support Centre?

Elizabeth Evans-Jones: I don’t recall which was first and which was second and which was third line.

Mr Stevens: But for second line support, let’s just deal with that, as a matter of generality first, from an IT perspective what does the second line support do, what is its purpose?

Elizabeth Evans-Jones: The purpose of second line support is to take an incident which can’t be resolved at the service desk, at first point of contact, and investigate further, attempt to resolve and, if resolution is not possible, then to pass that through to third line support.

Mr Stevens: In terms of – I’m not sure if you can say this as a matter of generality, but in terms of proportions of problems that should be resolved at first line or at second line or at third line, is there a general rule of thumb as to how many incidents should be capable of being resolved by first line support and then second line?

Elizabeth Evans-Jones: Speaking in general terms, no, it depends very much on the nature of the service that’s been provided, the access that a service desk may have. So, no, I don’t believe it’s possible to generalise to say how much should be resolved at first point of contact.

Mr Stevens: Please can we turn to your witness statement, and it’s paragraph 14, on page 4. You say that to explain your role:

“… I will briefly outline Fujitsu Core Services and Account model, as it was in existence when I was employed by the company. At this point in time, services in Fujitsu were either provided by Core Services or were Account Owned Services.”

Please could you explain what Fujitsu Core Services were?

Elizabeth Evans-Jones: Absolutely. I tried to articulate in the following paragraph, in paragraph 15. So Core Services were services that would be provided to multiple accounts. So the examples that I gave in paragraph 15 would be, for example, the service desk or engineering services, and the reason for that was that they were activities that could be customised for particular accounts. So a service desk has very much the same purpose for one account as for another account; engineering, again, very much the same purpose for one account or for another account.

So Fujitsu, at the time, had the model of having these Core Services accounts and then anything that was very specific to an account, for example, in Post Office, would be part of the account team. One of the key differentiators there is the fact that the resource and the management of those services resided with Core Services and the Core Services management structure, whereas any account owned services resided with the account for its management and its performance levels.

Mr Stevens: You’re quite right. You do say in your statement that the Helpdesk was a core service. So does that mean that the people who were dealing with Helpdesk enquiries, sitting on the phones, would also be dealing with calls related to different accounts?

Elizabeth Evans-Jones: No.

Mr Stevens: No?

Elizabeth Evans-Jones: So they were ring-fenced resources that were dedicated to Post Office, they were trained to support the Post Office Account. But the management structure was under Core Services. So the operations manager sat in Core Services, again ring-fenced for Post Office Accounts. There were other accounts sitting in Core Services where there were shared services but Post Office Account was not one of those. The resources were dedicated to Post Office Account or supporting Post Office.

Mr Stevens: On that point, please could we bring up FUJ00080478, page 8, please. This is a document concerning the Horizon Service Desk and described as a joint working document. If we could just go to the bottom of this page, please, just to get the date. It’s 4 September 2008, and if we could focus in on paragraph 2.4, please. It says:

“Fujitsu Services may provide a non-dedicated service desk function sharing the resource with other Fujitsu Services customers as described within this Service Desk, Service Description.”

So is it the case that Fujitsu may be entitled to have a service desk which dealt with other Fujitsu accounts?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: But from your time there and your recollection, it did not, in fact, do that?

Elizabeth Evans-Jones: Absolutely. Absolutely. I guess that’s the nature of Core Services. Shared services desks could be put in place but for Post Office Account it was a dedicated desk due to the size of the account.

Mr Stevens: Do you recall how many people were available to work in the Helpdesk whilst you were working there?

Elizabeth Evans-Jones: I don’t have the recollection as to how many people.

Mr Stevens: Do you recall the types of minimum qualifications that a person would need to be employed on the Horizon Helpdesk?

Elizabeth Evans-Jones: Again, I don’t recall the qualifications. I wasn’t involved in the selection of the resources to go on to the Service Desk. I can speculate that it was due to – it was the client engagement, the ability to communicate effectively with end callers, to be able to deal with sometimes challenging conversations, but – and IT experience, and again that would be my speculation based on my experience of running other service desks.

Mr Stevens: Would you – again I appreciate you don’t have knowledge of what these people actually required or what the qualifications were, but in terms of – from your experience, would you expect that people working on the Helpdesk would need some form of IT qualification?

Elizabeth Evans-Jones: Yes.

Mr Stevens: What level would that be?

Elizabeth Evans-Jones: Again, it depends on the service desk that’s being supported and the level of technicality of the service desk, but fundamental understanding of IT services would be, in my opinion, a requirement to be on an IT service desk.

Mr Stevens: Was there anything about this Helpdesk, the Horizon Helpdesk that took it out of the norm that meant more advanced qualifications were needed or less?

Elizabeth Evans-Jones: No. The opposite. The Horizon Service Desk, from my recollection, had very limited opportunity to resolve at the first point of contact. So, from my recollection, a lot of the calls that came through were related to hardware, a reboot was the maximum that the Service Desk could do there and that would be dispatched to engineers.

There was also a Knowledge Base that laid out step-by-step instructions as to what the Service Desk could do. But to my – the best of my recollection, the IT Service Desk, the Horizon Service Desk was not a technical service desk, not particularly technical.

Mr Stevens: I certainly want to come to explore some of those issues shortly. Before doing that, do you recall the training that was made available to members of the Horizon Helpdesk?

Elizabeth Evans-Jones: I don’t. I know that there was training and I refer to that in my statement. I know there was a training programme put in place. I don’t recall the duration nor do I recall the contents of that.

Mr Stevens: Do you know who would be responsible for ensuring that members of the Horizon Service Helpdesk – sorry, Horizon System Helpdesk, were properly trained?

Elizabeth Evans-Jones: The operations manager for the Horizon Service Desk in Core Services.

Mr Stevens: Who was that during your time there?

Elizabeth Evans-Jones: Paul Gardner, ^name I believe was his name.

Mr Stevens: Before moving on, if you can help us with this core or account – so Core Services or Account Services, do you remember whether the second line support, SMC, would be Core or Account Services?

Elizabeth Evans-Jones: I don’t recall. I don’t recall.

Mr Stevens: Let’s move to look at your role, then, as Service Delivery Team manager. We don’t need to bring it up but paragraph 17(a) you say that you managed the team of service delivery managers –

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: – who provided both Core Services and account owned services. How many people or service delivery managers did you manage?

Elizabeth Evans-Jones: I believe it was around eight or nine service managers. It changed over the two years. Eight or nine is my recollection.

Mr Stevens: How many of those would be responsible for work relevant to the Horizon Helpdesk?

Elizabeth Evans-Jones: I had one service delivery manager who was the key interface for the Horizon Service Desk.

Mr Stevens: Who was that?

Elizabeth Evans-Jones: I don’t recall his name, the role changed. Ian Mills, I believe, at one point was involved in the Horizon Service Desk. I don’t recall the name.

Mr Stevens: What was his day-to-day responsibility for the service desk?

Elizabeth Evans-Jones: So he would interface with – he would almost be the conduit between the account team and the Core Services team that provided the Horizon Service Desk. So he would – Ian or other people who held that role – would be looking at the metrics – the Service Desk metrics in terms of average speed of answer, dealing with any escalations that came through, making sure that the Desk was resourced appropriately. So he would work very closely within the operations manager for the Horizon Service Desk and also interact – he was the representation and interaction with Post Office, as well. We had operational reviews around the Horizon Service Desk performance.

Mr Stevens: Did the operations manager report to you?

Elizabeth Evans-Jones: Paul Gardner?

Mr Stevens: Paul Gardner?

Elizabeth Evans-Jones: No.

Mr Stevens: Who did Paul Gardner report to?

Elizabeth Evans-Jones: His management structure in Core Services.

Mr Stevens: Who did you report to?

Elizabeth Evans-Jones: The head of service delivery management for Post Office Account.

Mr Stevens: Let’s look, then, in more detail at the Helpdesk itself. Please can we look at your witness statement again, page 5, paragraph 17(a). You set out your role “To manage a team of SDMs” and, in respect of the Horizon Service Desk:

“… involved engaging with the Core Services Operations Manager to ensure delivery against the agreed performance metrics of the first-line desk, and improvement of the service. Ensuring that the Core Service function was in line with the profit and loss/business case, the [Horizon Service Desk Service Delivery Manager] also managed escalations from [Post Office Limited] on the performance of the Service Desk with the Core Services Team.”

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: So is it fair to say in terms of both – your responsibilities were both for the volumetrics, in the sense of how many calls were answered, the speed of the calls.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Was the quality of the advice provided also within your responsibility?

Elizabeth Evans-Jones: Yeah, so to clarify, the actual achievement of those performance metrics sat with the operations manager in Core Services, as did the quality. In the event that those metrics dropped down or the quality dropped down, that would then be discussed by myself and by the Horizon Service Desk SDM and we would work collaboratively with Core Services to implement improvements to address the quality metrics or to address the performance metrics as well.

And, as mentioned in my statement, if we received escalations from Post Office on the quality or on the performance metrics, we would work collaboratively to address those. So there were three parties involved in the process: the Core Services team, the account team and Post Office Limited.

Mr Stevens: Just to make sure I’ve got this, the operations manager, Paul Gardner, he had day-to-day responsibility for ensuring that the quality and volumetrics were to the required standard?

Elizabeth Evans-Jones: Yes.

Mr Stevens: You would monitor that –

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: – and step in when there was a drop –

Elizabeth Evans-Jones: Mm.

Mr Stevens: – and come up with ways or devise strategies to improve it to get it back to the contractually agreed level of service?

Elizabeth Evans-Jones: Yeah, there was almost the layers-off accountability. So Paul was accountable for the Service Desk, myself and my team were accountable to Post Office for those quality and the service metrics. We had to make ensure the service was delivered in line with the contractual metrics.

Mr Stevens: So in doing that role, whilst not immediately day-to-day responsible for the service desk, you had awareness of its operation and its function?

Elizabeth Evans-Jones: Absolutely, yeah.

Mr Stevens: Turning then to its function, you state in your witness statement, you say that – this is paragraph 20:

“The HSD was primarily a log and flog function, as there were very limited first line/level fixes that the desk could complete.”

Elizabeth Evans-Jones: Yeah.

Mr Stevens: Can I ask you to expand on “log and flog”?

Elizabeth Evans-Jones: Yeah, so as referred to, the Horizon Service Desk really had limited opportunity to resolve at first point of contact. So log and flog is a generic term used in the industry which is basically to log a ticket and then pass it through to the next level of support, be that hardware, software or query management.

Mr Stevens: Why were there such limited first line fixes available for the Horizon Service Desk?

Elizabeth Evans-Jones: So a large number of the incidents that were logged were hardware related. Again, as I’ve articulated, very little could be done on a hardware issue, apart from to try and reboot the counter. If the counter was down, that caused issues for the Post Office. So the approach that was taken if the reboot didn’t work – and that normally took about 20 minutes – we’d dispatch the engineer to get an engineer on site as quickly as possible to allow the branch to trade again.

Single-counter branches was obviously more critical than multi-counter branches. Again, with keypads – and, again from my understanding, there was very limited software fixes that the desk could do anyway because I don’t believe they had access to fix anything with the software. They’d look in the Knowledge Base, if there was no immediate resolution that was documented in that, they would then pass that through to the second line support or third line support.

Mr Stevens: Again, something else that we will come to in due course but the types of calls, let’s just cover that for a moment. You said there were a lot of hardware calls?

Elizabeth Evans-Jones: Mm-hm, yeah.

Mr Stevens: The Inquiry has heard a significant amount of evidence from subpostmasters who stated that they faced discrepancies in their accounts, which were generated by Horizon. Do you recall there being a significant number of calls relating to discrepancies which came in to the helpdesk?

Elizabeth Evans-Jones: I would only have the classification of which the ticket was logged against, the PowerHelp codes. I don’t know – from best of my recollection, I’m not sure there was a code that specifically called out discrepancy. So I don’t know. From my recollection, the bulk of the calls that came through were hardware related calls or branch services were offline, as in the BT network that was put in place to the Post Office was offline, which meant the branch couldn’t trade.

Mr Stevens: Please can we just bring up paragraph 41 of your witness statement at page 16. Thank you. Here you say that:

“As a result of some of the escalated incidents, which I directed to the Software Team, I was aware that it had been reported by SPMs that the system could cause branch discrepancies, however I do not recall these in detail.”

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Can you just summarise when you would become involved in these escalated incidents, in your role?

Elizabeth Evans-Jones: Yeah, escalations generally came from two sources, one was Post Office would escalate to me directly or the Service Desk would escalate to the Service Delivery Manager for that function, and then they would escalate to me if they were unable to resolve that escalation.

Mr Stevens: You say there you were aware of the reports of SPMs saying that the system could cause discrepancies.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Was that common knowledge in the Helpdesk, of the fact that SPMs were making such allegations?

Elizabeth Evans-Jones: I’m unable to comment on whether the Service Desk thought this was common knowledge or not. As I say, the – I really have no recollection of whether I knew that or not.

Mr Stevens: Stepping back, then, from the Service Desk, amongst your colleagues you worked with day to day, was it a known fact that allegations by SPMs were being made that the Horizon System could cause discrepancies?

Elizabeth Evans-Jones: Again, it’s not an area that I was particularly involved in. I think there was some awareness that there were discrepancies but, again, I’m not sure of how widely that was known, nor did I have any understanding about the scale of discrepancies that could be caused.

Mr Stevens: Are you aware of anything that was done to investigate those allegations within Fujitsu?

Elizabeth Evans-Jones: Not personally. As I say, my portfolio wasn’t around the software. It’s my speculation that it was being investigated by the second and third line support teams in the software side of the support.

Mr Stevens: I want to look at one of those escalations now, and turn to a document which you were given this morning. It’s POL00028984. If we could go to page 10, please, at the bottom. This is an email that the Inquiry has seen before. It’s from Sandra MacKay to Shaun Turner. It says:

“Shaun, you may recall that in September the above office had major problems with their Horizon system relating to transfers between stock units.”

Go over the page, please. Thank you.

The SPMR has reported that he is again experiencing problems with transfers, (5 January ‘06) which resulted in a loss of around [£43,000] which has subsequently rectified itself. I know the SPMR has reported this to Horizon Support, who have come back to him stating that they cannot find any problem.”

If we could go then to page 8, please, and to the bottom. There’s an email there from Gary Blackburn to you on 15 February 2006. Do you recall Gary Blackburn?

Elizabeth Evans-Jones: I do, yeah.

Mr Stevens: Who was he?

Elizabeth Evans-Jones: He worked on Post Office. I can’t recall his exact role but he was one of three or four people that I had regular contact with on the Post Office Account.

Mr Stevens: He forwards this email on to you –

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: – describing the detail, and says:

“… could you please update me on the corrective action plan as this still appears to be occurring within the branch.”

Do you have any recollection of this matter?

Elizabeth Evans-Jones: No, the first I remember of this is when this document was presented to me this morning.

Mr Stevens: If we go above, your response is – just move up slightly so we can see the date, please. Thank you.

On 16 February, you say:

“Hi Gary,

“I have checked the call and this issue is scheduled to be resolved in S90.”

Do you recall what S90 was?

Elizabeth Evans-Jones: I don’t.

Mr Stevens: If I said it was a release, a software release?

Elizabeth Evans-Jones: I assume it’s a release, yeah. I don’t recall specifically what it is, no.

Mr Stevens: I appreciate you say you can’t recall this incident at all, but how would you have gained this information to come back to Gary Blackburn with this response?

Elizabeth Evans-Jones: I would have spoken to the second or third line support, third line in this case, for the Software Support Team.

Mr Stevens: Do you think you would have been concerned that the Helpdesk had initially advised the postmaster that this discrepancy was not a fault in the system when it transpired that it was a software problem?

Elizabeth Evans-Jones: Yes, I believe I would have been concerned. The Horizon Service Desk would have followed whatever was in the knowledge database, so I would have been concerned that the information in the knowledge database would have been incorrect and that incorrect advice would have been given to the subpostmaster or the postmaster.

Mr Stevens: Can you recall if any steps were taken to address that concern?

Elizabeth Evans-Jones: The KELs, the knowledge articles, were updated on a regular basis. They weren’t updated by the Service Desk, again, so I can’t recall if anything specifically happened in relation to this particular incident. However, there was a process to make sure that the KELs were updated with the latest information.

Mr Stevens: When KELs were updated like that, was it simply a case of there’s a KEL on the system, an update has been made, so the next time someone accesses that KEL, they will see updated information, or was there a circular sent round to members of the Helpdesk to advise them of any updates?

Elizabeth Evans-Jones: Um, again, I don’t recall in detail. I know the KELs would have been updated and that information, the latest information, would have been visible. To the best of my recollection, there was a process whereby information was circulated round the Service Desk but that wasn’t for every single KEL that was updated.

Mr Stevens: Do you have any recollection of which KELs would be –

Elizabeth Evans-Jones: I’m afraid I don’t, no.

Mr Stevens: At page 5 of the document, if we can go down slightly please to the bottom – thank you – we see Gary Blackburn emails you back on 17 February. He’s got some questions, which are over the page.

In particular, one of them is:

“Do we understand why this particular branch has been having problems? Or are there other branches in the network that have been having this problem?”

If we go back to, I think, page 5, you send that on to Mike Stewart. Do you remember who Mike Stewart was?

Elizabeth Evans-Jones: Yeah, he was a Service Delivery Manager who reported to myself and worked on online services.

Mr Stevens: What was the purpose of sending this to him?

Elizabeth Evans-Jones: So he was closer to the applications and the systems to be able to investigate that. So it was common that I would then distribute the work to the people who had more knowledge around the content of the email.

Mr Stevens: Do you recall, after sending this email, if you had any more involvement with this issue?

Elizabeth Evans-Jones: I don’t recall the email. So, no, I don’t believe I had any further involvement. From what I can see from the emails, I was even taken off the email exchange.

Mr Stevens: Can we go to page 3, please, and the email from Anne Chambers to Mike Stewart on 23 February. This isn’t an email – well, there’s no evidence here to suggest you were sent this at the time. I just want to look at the second paragraph though, which says:

“Haven’t looked at the recent evidence, but I know in the past this site had hit this Riposte lock problem 2 or 3 times within a few weeks. This problem has been around for years and affects a number of site most weeks, and finally Escher say they have done something about it.”

So this is, is it fair to say, talking about a fairly significant bug in the Horizon System code?

Elizabeth Evans-Jones: It appears that way, yes.

Mr Stevens: Were you aware of this at the time at all?

Elizabeth Evans-Jones: As I say, the only recollection I have now is from this email that was sent to me, was provided to me this morning. Until this point in time, I’d no recollection of the Callendar Square issue nor this Riposte lock problem.

Mr Stevens: If this information had been given to you at the time, do you think it’s something that you would have remembered now?

Elizabeth Evans-Jones: Absolutely. Because, you know, it’s a significant issue and I would have absolutely done to the best of my ability to make sure we investigated that properly. So yeah and that’s because of the person I am. So …

Mr Stevens: Thank you. That document can come down now.

We’ll move on to a different matter which is the types of calls you were referring to earlier, and if we could bring up FUJ00083429. This is a Fujitsu Services Post Office Account Service Review Book for February 2007. Can you briefly summarise what this document was or the purpose of these types of document.

Elizabeth Evans-Jones: Yeah, it was a contractual obligation that each month the Fujitsu Post Office Account had to provide this through to Post Office and it outlined the performance metrics for the key services that Fujitsu provided. So there were performance metrics and commentary included in there.

Mr Stevens: Please can we turn to page 11. This is showing the Horizon Service Desk, the table, unhelpfully in black and white, but we can come to the numbers further down, but is this showing that – essentially showing the metrics for from February ‘06 to February 2007 –

Elizabeth Evans-Jones: Yeah.

Mr Stevens: – for the service level agreements?

Elizabeth Evans-Jones: No, this is showing the number of calls in each of those categories, so the number of calls was not the service level agreement. The service level agreement was more around average speed of answer. Number of calls that went through to voicemail, for example, I believe those are listed in one of the statement of work documents.

Mr Stevens: If you could go to the bottom of this page, please, and if we could make the table at the bottom just slightly bigger – thank you.

So we see the total calls, third up from the bottom, and a monthly call limit.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: What was the monthly call limit?

Elizabeth Evans-Jones: I don’t recall specifically. I could speculate that that’s the document – that was a contractual level was put into a document, so that if we – if the number of calls exceeded or significantly were less than this, it would trigger a conversation with Post Office around the volumetrics of the service desk and the cost of the service. It’s standard for IT to have those threshold limits in there.

Mr Stevens: We see that the calls range in February ‘06 is just over 13,000, to this 16,000 in January ‘07, 15,500 in February ‘07.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: In terms of the breakdown of different types of calls in, significant numbers for hardware?

Elizabeth Evans-Jones: Yeah.

Mr Stevens: At the bottom, there is a collection for – or a category for software, as well. Can you recall or where – we discussed discrepancies earlier. Which category do you think discrepancies would fall into, reported discrepancies?

Elizabeth Evans-Jones: My assumption is that they would fall under software.

Mr Stevens: Thank you. That document can come down. If we could bring up FUJ00001966. This is a document dated 19 August 2005. So “Service Level Targets for Horizon Services”.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: So drafted just before, I think, you started on the Post Office Account?

Elizabeth Evans-Jones: Yeah.

Mr Stevens: At page 8, please, we have the service level targets for, at the bottom, Horizon System Helpdesk.

Elizabeth Evans-Jones: Yes.

Mr Stevens: The first three are, I think, to do with calls answered and the proportion there?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: We then have “Level 1 calls resolved within 5 minutes”, 95 per cent; do you recall what a level 1 call was?

Elizabeth Evans-Jones: I don’t recall exactly what a level 1 call was, however it was something that would be able to be resolved at the Service Desk.

Mr Stevens: The same for level 2, really. Do you recall the difference between that and a level 2 call?

Elizabeth Evans-Jones: So, again, a level 2 call, again, I don’t recall. So that would be something that – within the Service Desk, there was a level 1 and a level 2 sort of level of service. So level 2 probably had more time. They took more time to see if they could resolve at first point of contact. Obviously, it’s much more advantageous for post offices and for Fujitsu to resolve at the Service Desk rather than pass to a second or third line support team.

So seeing this now has triggered the memory in me that there was a level 1 and level 2 Service Desk within the Horizon Service Desk. I don’t recall the difference between a level 1 and level 2 call.

Mr Stevens: So this is saying of level 1 calls, say, for example, 95 per cent should be resolved within five minutes and for level 2 calls, those defined as level 2, 95 per cent should be resolved within 30 minutes. Can you recall if there were any targets or guidelines of the proportion of calls which come into the Helpdesk which should be level 1, which should be level 2 or which should be level 3?

Elizabeth Evans-Jones: No, I’ve no recollection of that.

Mr Stevens: That document can come down, thank you. That monitors how the Helpdesk in terms of volumetrics responded. How was the quality of the Helpdesk advice actually given, how was that monitored?

Elizabeth Evans-Jones: So there was – I believe they were called service controllers or the team leaders would monitor and listen into calls. That was then given feedback specifically to agents. Again, I don’t recall the percentage of calls that were listened in to but that was part of general Service Desk practice. There was also a complaints process. So when we received a complaint or the Horizon Service Desk would receive a complaint, that would then be logged and that would be investigated to determine if the complaint was a valid complaint or not.

Mr Stevens: Let’s move to that assistance and how that was given. Can we bring up FUJ00079939. This is a “[Post Office Account] Customer Service Incident Management Process” definition, drafted on 23 March 2005. This is for what we’ve called Legacy Horizon, or what’s known as Legacy Horizon, the version of Horizon in place from national rollout until 2010.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: You drafted the Horizon Online version of this document; is that right?

Elizabeth Evans-Jones: Yeah, that’s correct.

Mr Stevens: But this is the document that you would have been working with or would have been used at the time that you were in post?

Elizabeth Evans-Jones: Yeah, this was drafted just before I joined but, yes, this is the process that we would have been working to.

Mr Stevens: If we could turn to page 8, please. The “Process Objective”, under 1.2 says:

“The objective of this document is to define the process for Incident Management of the POA environment. For the purpose of this document an Incident is defined as:

“‘Any event which is not part of the standard operation of a service and which causes, or may cause, an interruption to, or a reduction in, the quality of that service’.”

So if a subpostmaster called the Helpdesk with a possible software problem, that’s an incident to be managed under this process; is that right?

Elizabeth Evans-Jones: That’s correct, yes.

Mr Stevens: Can we turn to page 12, please. So we have a flowchart here showing at the top, entry ways into the contact received by the POA Service Desk. “SDU”, is that Service Delivery Unit?

Elizabeth Evans-Jones: Correct, yeah.

Mr Stevens: In lay terms, what would a Service Delivery Unit be?

Elizabeth Evans-Jones: So the software support, the SSC was a Service Delivery Unit. So it was a team, a resolving unit, if you like, a resolving team that would work to resolve incidents.

Mr Stevens: So a team within Fujitsu such as the SSC or the –

Elizabeth Evans-Jones: Yes, or engineering, yeah.

Mr Stevens: There’s then “User”, which is presumably the subpostmaster?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: “System” and “Service Management”. Do you know what those are?

Elizabeth Evans-Jones: So we could have system driven alerts that would come up in. From my recollection they would come from the data centres, if there were any system alerts that could trigger an incident being logged at the Service Desk, and service management would be myself, my team and the wider service delivery organisation. Users would also be Post Office Limited, as well, so not just necessarily just subpostmasters, just for clarification.

Mr Stevens: No, and we don’t need to go through all of this flowchart. It starts with trying to triage the query, basically. At the bottom we see, if we just move down slightly, four types of outcome. There’s “Incident”, which then follows this process in this document; “Advise & Guidance, Answer enquiry and close or refer to … NBSC”; “Out of Scope”, that’s where – it wasn’t within the scope of the services provided by Fujitsu?

Elizabeth Evans-Jones: Correct.

Mr Stevens: And “Quality”.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Is “Quality” looking at the quality of service provided by the Helpdesk rather than the quality of the Horizon System?

Elizabeth Evans-Jones: From my understanding of the document, yes.

Mr Stevens: On the helpdesk, how would the – what process was used or guidance was used for allow a Helpdesk operator to decide whether it’s an incident or something that needs to go to the NBSC?

Elizabeth Evans-Jones: So the service desk would have had call scripts that they would go through and that would help them then determine which of these four categories the incident would be logged – or not necessarily the incident, which of the four categories would be applicable in this process.

Mr Stevens: I’m just going to move forward, actually, because you’ve mentioned call scripts now and I think in your statement you say that the agents were provided with scripts, pre-defined questions, which they were expected to use when providing support to the SPM.

Do you recall who was responsible for drafting those scripts?

Elizabeth Evans-Jones: The Service Desk team would have been responsible for drafting those scripts, however they would have had input from the Service Delivery Units or from service delivery management as well.

Mr Stevens: Where were they held for the operators to access?

Elizabeth Evans-Jones: I don’t know where they were held. I don’t know where that documentation was held. I don’t recall if it was actually part of the PowerHelp tool set. In some service management tool sets, the script is actually in the software, so it prompts the agent what to say. But, in this instance, I don’t recall where it was held.

Mr Stevens: Do you recall if there was a general script to follow for all calls or if there were individual scripts for specific issues raised by subpostmasters?

Elizabeth Evans-Jones: Again, I wasn’t actively involved in the day-to-day operation of the desk. My recollection is that there was a script that initiated the conversation, you know, greeting the caller, getting the Post Office branch ID – again, I can’t remember the correct terminology for that – and then obviously trying to capture specific information. I think one of the documentations has that in there, one of the joint working documents actually lists out some of the scripts that needs to be said.

Mr Stevens: I may have the document in my mind, which you’re referring to. Shall we bring up FUJ00080478. This is a Horizon Service Desk joint working document, and we see at the bottom the – you were an author on this with John Casey.

Elizabeth Evans-Jones: Yes, so John was one of the service desk managers reporting into Paul Gardner.

Mr Stevens: Please can we turn to page 13. This section looks at the end-to-end incident management and if we go down slightly, under 4.4.1, the third paragraph down, it says:

“The moment an Agent receives an incoming telephone call, they will greet the customer with the example shown below. All spoken words are marked in Italics and ‘quotes’.

“‘Good Morning/Afternoon/Evening. Horizon System Helpdesk [AGENT NAME] speaking.

“‘May I take your Branch code please?’”

Elizabeth Evans-Jones: Yes, and this would have been part of the training that was given to the Service Desk agent before they took live calls.

Mr Stevens: Are you aware if this document was converted into a more precise script that would then be used by the Helpdesk?

Elizabeth Evans-Jones: I am not personally aware but it would be my speculation that it was. This is, you know …

Mr Stevens: Please could we bring up now FUJ00138733. I think this was a document you were given this morning.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: It’s:

“PROCESS – ID:408501 – Engineer Refused Access Process.


“Please use this KA …”

Do you know what “KA” stands for?

Elizabeth Evans-Jones: I don’t know, no. Knowledge article, perhaps I would hazard a guess at.

Mr Stevens: Knowledge article. That’s what I was about to suggest.

Elizabeth Evans-Jones: Yeah.

Mr Stevens: But:

“… if an engineer has called to advise they have been refused access at a post office.


“If an engineer has called to advise that they have been refused access at a post office, please follow the below:

“Frontline Process …”

Then 2 says:

“HSD contacts site and follows call script below in purple.”

We don’t then need to read it out but there is then a call script there for a specific incident, in this case engineer refused access process. Do you recall seeing items like this during your time working with Fujitsu?

Elizabeth Evans-Jones: No, I don’t recall seeing this. Again, I can speculate that this is the knowledge article that the Service Desk had, that would – they would refer to this in the event that they received a phone call from an engineer saying they didn’t have access to the site.

Mr Stevens: Would you anticipate that there would be similar articles for – this is obviously engineer refused access but say a subpostmaster rang with a discrepancy would you expect there to be scripts of a similar nature advising the operator how to deal with that?

Elizabeth Evans-Jones: It would be my expectation, yes.

Mr Stevens: That document can come down. Thank you.

Do you recall ever an instruction being given to helpdesk staff to tell subpostmasters that they were the only person experiencing a problem that they had reported?

Elizabeth Evans-Jones: No.

Mr Stevens: Would you expect – what would you say if such advice was given?

Elizabeth Evans-Jones: I would say that would be erroneous advice.

Mr Stevens: If we could please bring back up FUJ00079939. If we could go to page 15, please, paragraph 2.4. This sets out, I think it’s fair to say, what the Service Desk was expected to do when handling calls or incidents and, for the record, it says:

“The Service Desk agent then attempts to resolve the Incident using the resources available. This starts by interrogating HSH ONE …”

Do you recall what that was?

Elizabeth Evans-Jones: I don’t recall. Again, I can speculate that that was the term given to the system that housed the KELs or it was a precursor to the knowledge database. That’s a terminology that I don’t recall.

Mr Stevens: “… to find all information related to the Incident symptoms. If the Incident is routine, ie there is a pre-determined route for resolution, then the Incident is referred to the relevant SDU using the Service Desk Support Matrix in HSH ONE.”

When you say “SDU”, that would be perhaps the SMC or engineering?

Elizabeth Evans-Jones: Or engineering, yeah.

Mr Stevens: It then goes on to say:

“If the Incident is not routine, the Service Desk agent checks for Known Errors listed in HSH ONE and the SSC KEL against records relating to the Incident symptoms. If a match is found, the agent informs the caller of the workaround or resolution available and links the call to the master Incident record.”

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Do you recall the KEL database?

Elizabeth Evans-Jones: Not in detail, I recall its existence.

Mr Stevens: Do you recall whether members of the helpdesk found that an easy system to use?

Elizabeth Evans-Jones: I wouldn’t be able to comment on that. I don’t know.

Mr Stevens: During your time analysing the call metrics, et cetera, would you have needed to consider, for example, if there was an increase in delays in resolving calls within 10 minutes or 30 minutes, would that be something you would look into, the reasons for the delay, sorry?

Elizabeth Evans-Jones: Yeah, for the level 1, level 2 on the Service Desk. So anything that got routed to a Service Delivery Unit, the only one that I would have had any involvement in is the engineering service and anything that was rooted to Cable & Wireless or BT for the online branch services – it wasn’t Internet – ADSL as it was at the time, but I wouldn’t look into any of the software calls.

Mr Stevens: Who would look into the software calls?

Elizabeth Evans-Jones: The SSC would be my assumption.

Mr Stevens: So let me look – put this a different way. When examining whether a software – a number of software calls had been resolved quickly enough and within service level targets, would you ever have looked into whether the KEL database was an effective way or was effective in giving Horizon Service Helpdesk operators information they needed to resolve level 1 and level 2 calls?

Elizabeth Evans-Jones: The only metrics that I’d have looked into were the level 1, level 2 within the Horizon Service Desk. So the – I don’t know what the resolution timescales or the SLAs were expected to be for the Service Delivery Units that were not part of my portfolio. So I don’t know what the SSC SLA was or what it was intended to be. But going to the question, looking at level 1, level 2 within the Horizon Service Desk, if we saw deterioration or that service level metric wasn’t being met we would look to try to understand what was the root cause of that.

Mr Stevens: Sir, I don’t have much longer to go but, for the transcriber, I notice we’ve been an hour, so I wonder if we could have a short break?

Sir Wyn Williams: Yes, by all means. Where are we now? 11.30 all right?

Mr Stevens: Yes, that’s fine, thank you, sir.

Sir Wyn Williams: Good, see you then.

Mr Stevens: Thank you.

(11.13 am)

(A short break)

(11.30 am)

Mr Stevens: Sir, can you see and hear me?

Sir Wyn Williams: Yes, I can, thank you.

Mr Stevens: I want to go back to the document we were on and at the same place, please. It’s FUJ00079939. And paragraph 2.4 – sorry, 2.5. We went to this paragraph beforehand and this is where the Service Desk operator couldn’t resolve the problem of using HSH One –

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: – and then checked for known errors listed in the same database but also in the SSC KEL database that we discussed?

Elizabeth Evans-Jones: Yeah.

Mr Stevens: It said:

“If a match is found …”

So presumably if it matches to something in the One system or in the SSC KEL system:

“… the agent informs the caller of the workaround or resolution available and links the call to the master Incident record.”

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Can you just explain what the master incident record was in that case?

Elizabeth Evans-Jones: So it’s common practice in the Service Desk to create a master incident record and then append what we call child incidents to that master record. That then allows any Service Desk or, in this case, the Post Office Service Desk, to be able to capture the metrics on how many occurrences of that incident there actually were, if that makes sense. So it’s a mechanism of saying we have this major incident, and then there’s appended other incidents beneath that.

Mr Stevens: So for instance with Callendar Square – I’m not saying this is what happened but to use it as an example, there may be a master bug or incident and then each time one is identified in the field, in theory, that should be appended to that incident as a child?

Elizabeth Evans-Jones: That’s the theory, yes.

Mr Stevens: Is that different to the KEL, so would the KEL be updated to show that the call had been raised and was linked to the overall KEL?

Elizabeth Evans-Jones: There would be a cross-reference between the KEL number and the number of incidents. So, again, this is my assumption, that the master incident record would refer to the KEL and, therefore, you would be able to extrapolate that KEL12345 had X number of incidents associated with it.

Mr Stevens: So, in this case, if there was a second incident of a similar or the same materialisation of a bug, we had the child. You think that the – there’s a link to the actual KEL, so on the KEL you can see incidents linked to it. Are you aware if that ever changed?

Elizabeth Evans-Jones: So just to clarify, I’m not sure that the KEL would show the number of incidents but you could cross-reference the two data sources to achieve the same outcome, so just for clarity – and sorry, what was the second question?

Mr Stevens: That system that you’ve just referred to, are you aware if that ever changed during your time at Fujitsu?

Elizabeth Evans-Jones: I don’t believe so, no.

Mr Stevens: Thank you. Can we turn to page 16, please, and paragraph 2.6:

“If there is no match in HSH ONE or the SSC KEL, the Problem Database is checked for current incidents outstanding. If a match is made, the caller is then advised of the status of the Problem and the call is then linked to the master Incident record given in the problem details.

“2.7. If no match is made against the Problem Database, the Service Desk continues with first line resolution of the Incident assisted by the Product Support Engineers …”

Who were the product support engineers?

Elizabeth Evans-Jones: I don’t recall who the product support engineers were in this particular instance. Again, I could speculate that they are subject matter experts associated with the particular software or hardware but I don’t recall specifically who the PSEs were in this particular instance. As it’s articulated here, it appears that they’re part of the Service Desk.

Mr Stevens: So and then 2.8:

“If the PSEs cannot resolve the Incident, it is referred to the relevant SDU using the Service Desk Support Matrix in HSH ONE.”

Elizabeth Evans-Jones: Yes.

Mr Stevens: So is my understanding right that this, you follow this process and then if this doesn’t lead to a resolution, it’s then passed to second line support?

Elizabeth Evans-Jones: So what would happen is, if an incident is logged – and this is sort of generic service management best practice – if an incident is logged, it’s then validated to see if there’s a known solution, a known workaround to get the service restored. That’s the KEL that would be looked in. The KEL contained – or the knowledge database contains how to resolve an incident, how to restore that incident.

The problem database would be open items for which the resolution hasn’t yet been identified. So problem management is one level elevated to incident management and then the product support engineers, reading this now, they are subject matter experts in the Service Desk and if they are unable to resolve, that’s then when it would get that then passed through to the Service Delivery Unit, should that answer your question or not.

Mr Stevens: I suppose my question is that’s when it goes to second line?

Elizabeth Evans-Jones: Mm-hm, yeah.

Mr Stevens: So the Helpdesk – in your experience, did the Helpdesk follow this as a matter of course, in practice, this system?

Elizabeth Evans-Jones: Yes, yeah, absolutely. That would be the system that they followed. I can’t say that they followed it 100 per cent of the time because there are humans involved in this but that was absolutely the intent: to check the KEL, to see if there was a resolution and then to check the problem database and, if it was unable to resolve, to assign it through to the SDU. The KEL might actually stipulate in it “You need to pass it to the Service Delivery Unit”, so that could also be some of the information that’s in the KEL.

Mr Stevens: Your evidence earlier when we discussed the log and flog matter, you discussed how there weren’t many first line fixes available.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: When we explored that, you referred to the number of hardware issues –

Elizabeth Evans-Jones: Yes.

Mr Stevens: – and we saw the number of calls that came in to do with hardware, so those calls would be, of course, passed straight on to the engineering department or whatever it is.

Elizabeth Evans-Jones: Yes.

Mr Stevens: Limiting it to software complaints, was there still a limited number of first line fix available to the Helpdesk when this process was followed?

Elizabeth Evans-Jones: It’s my belief that, yes, there would still only be a limited number of fixes that the Service Desk would be able to do.

Mr Stevens: Do you know why that was?

Elizabeth Evans-Jones: Again, I – my speculation is that they didn’t have any ability to resolve software incidents and they had to go to the specialised teams. It’s very few service desks that are able to resolve software issues. Mostly it’s a reboot to see if that solves it but that’s why it’s my belief that there was very limited first-level fix that the Service Desk could do in software incidents.

Mr Stevens: Were the types of software incidents that were being presented to the Helpdesk more complex than you would expect in other IT projects?

Elizabeth Evans-Jones: I’m unable to comment on that. Again, the information I saw was the breakdown of the number of tickets logged against specific PowerHelp codes.

Mr Stevens: If we go to a different document, please. It’s FUJ00079897. It’s a 2003 document, “End-to-End Support Process, Operational Level Agreement”.

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Please can we turn to page 6. This sets out “HSH/HIT/SMC obligations to SSC”. I think we’ve covered all of those abbreviations, save for “HIT”. Do you remember what that was?

Elizabeth Evans-Jones: Well, according to the abbreviation definition in the document it’s the Horizon Service Desk Incident Team.

Mr Stevens: Do you recall what their role was?

Elizabeth Evans-Jones: There was a subsection within the Service Desk – the Service Desk have – it’s not just agents that have responded to calls. You also have team managers, PSEs, as I now remember, and incident management teams. And the incident management teams would look at major incidents or significant incidents and make sure that those were passed through to either the service management team or managed in accordance with the incident management process – the major incident management process.

Mr Stevens: So this document is looking at what is described as obligations for first and second line support collectively to third line support?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Subparagraph (d), it says:

“To ‘filter’ all calls for which the problem is already known to the support community and for which a resolution is already known or has been generated. In this context the term ‘resolution’ can take a number of forms, including.

“The statement that the problem is resolved in release [X] of the Horizon solution.

“There is a documented workaround for the problem.

“The documentation relating to that part of the system is under review or being changed.

“No calls passed to the SSC which are subsequently resolved as known errors, except in cases where the symptoms as reported by the customer did not match the symptoms recorded in the known error documentation, and which therefore the HSH/HIT/SMC could not reasonably have been expected to find.”

So this is essentially putting into practice what you described earlier, that, where possible, first and second line support should resolve the calls where they can do?

Elizabeth Evans-Jones: Mm-hm.

Mr Stevens: Are you aware if there was any consequences of or what happened when the SSC considered that a call had been directed up to them inappropriately?

Elizabeth Evans-Jones: Um, they – my recollection is that they would refer that back and we would try and have a closed loop process to understand why a call had been passed to SMC which – or SSC, sorry, which shouldn’t have actually been passed there, because the intent is always to try to resolve as quickly as possible. So that’s a failure in the process, if something has gone through to a third line support team, which should have been resolved or could have been resolved by a level 1 or a level 2 desk.

Mr Stevens: To what extent was there pressure on people in the Helpdesk to resolve calls themselves rather than refer them up?

Elizabeth Evans-Jones: Um, again, I wasn’t actively on the – involved in the day-to-day operation on the Service Desk. There was, you know, there was a requirement for them to follow the process correctly but I wouldn’t say that there was pressure on them to not pass calls through to second or third line. You know, there was no metrics on that and the desk was operating on its – on the metrics that we discussed in the previous documentation. So I don’t believe there was undue pressure or any pressure for them to not refer calls inappropriately.

Mr Stevens: Can we turn to page – I think it’s just over the page, subparagraph (m). Just further down, please. We have (m), which is:

“To ‘filter’ all user error calls and ensure that they are closed.

“No calls passed to SSC which are subsequently closed as ‘user error’.”

Then (o):

“To ‘filter’ all calls for which the Pathway software [it says ‘in’ but ‘is’] not at fault.

“No calls passed to SSC which are subsequently closed as ‘No fault in product’.”

From the Helpdesk perspective, do you think the people working on there had sufficient expertise to be able to determine whether a call was – or an incident was caused by user error, rather than the software itself?

Elizabeth Evans-Jones: The intent of the knowledge articles is to provide the knowledge to the Service Desk agent, so that they should follow the script that’s in the knowledge article and that would then determine whether it was – how to route the call. The intent of knowledge articles is to eliminate that need for in-depth knowledge for Service Desk agents. So I don’t believe, if the knowledge article was written correctly, then they should have been able to follow that and that would have then delivered the right – the correct outcome.

Mr Stevens: That document can come down. Thank you.

I’ve been asked to ask you if whether, to your knowledge, there were members of the helpdesk who were ever advised to tell subpostmasters to accept discrepancies because they were caused by user error.

Elizabeth Evans-Jones: Absolutely not to my knowledge did that happen.

Mr Stevens: Finally, please could we bring up statement WITN06660100. This is a witness statement from Amandeep Singh, who will be giving evidence to the Inquiry later today and worked at the Helpdesk before your time at Fujitsu in 2001, in Wakefield. Can I ask you to turn to page 3, please, of the statement. I’ll just read it for the record. It says:

“The floor on these days …”

When it says “these days”, it is referring to Wednesday when there was balancing issues:

“The floor on these days was most toxic with vocal characters in Squad A, unchallenged by managers who looked away as all Asians were called Patels, regardless of surname. Shouts across the floor could be heard saying ‘I have another Patel scamming again’. They mistrusted every Asian Postmaster. They mocked Scottish and Welsh postmasters and pretended they could not understand them. They created a picture of postmasters that suggested they were incompetent or fraudsters.”

Were you aware of any such behaviour on the Helpdesk during your time at Fujitsu?

Elizabeth Evans-Jones: No, not at all. And reading that, I find that absolutely appalling.

Mr Stevens: I have no further questions, but before I ask if the Core Participants have questions, is there anything further you would like to say to the Inquiry?

Elizabeth Evans-Jones: No, that’s fine. Thank you.

Mr Stevens: Yes, Mr Stein has some questions, sir.

Questioned by Mr Stein

Mr Stein: Sir, one area of questioning, it won’t take long.

Ms Evans-Jones, I represent a very large number of subpostmasters and mistresses, all of them have been affected by this scandal. Dealing with your knowledge of the support systems, can you help me whether the first line support groups used the same incident logging system as the rest of the support chain?

Elizabeth Evans-Jones: Um –

Mr Stein: Now, first of all, do you want me to repeat the question?

Elizabeth Evans-Jones: Can you define “support chain”? Are we talking SDUs?

Mr Stein: Yes, well, I am quoting, in fact, from a document. It is the document after your time, relating to these matters. So what I’m trying to find out is whether the original Horizon System had the same problem. So all I’ve got is that the first line support groups – so I imagine are the helpline support providers. So if we look at it from that perspective, did they, in your time, use the same incident logging system as the rest of the support chain, which would then be the lines 2, 3 and 4?

Elizabeth Evans-Jones: To the best of my knowledge, the Service Desk used PowerHelp initially. That then changed to TRIOLE for Services. PowerHelp, from an engineering perspective, was not the system used by the engineering and it transferred into a Core Services tool set that managed the engineering and, to the best of my recollection, SSC from the software perspective, had access to the PowerHelp that they transferred it into their own tool or that they worked on.

Mr Stein: So is the answer to my question that they didn’t, in fact, use –

Elizabeth Evans-Jones: To the best of my knowledge, I think different systems were used.

Mr Stein: Yes. Thank you.

Sir Wyn Williams: Anyone else?

Ms Page: I do have some questions, please, sir.

Sir Wyn Williams: Yes.

Questioned by Ms Page

Ms Page: It’s Flora Page, also representing a number of subpostmasters. Can I ask, please, for document FUJ00120049 to come up, please, and if we can go to page 6. If we can go to the definition of – I understand this to be something which would deal with problems which then go into what you’ve described as the problem database; is that right?

Elizabeth Evans-Jones: Yeah, yeah. So for clarity, a problem is an issue that doesn’t have a documented workaround or resolution, so an incident, and then you move into problem management, and then change management addresses the root cause that’s in the problem, the kind of three flow through to each other.

Ms Page: Well, as I understand it from this document, the relationship between an incident, which we’ve already seen the definition of, and a problem, is that the problem is defined as – let’s see if I can find it, it’s that second sentence of the first paragraph there:

“For the purpose of this document a Problem [with a capital P] is defined as the unknown underlying root cause of one or more Incidents.”

Elizabeth Evans-Jones: Mm-hm.

Ms Page: Then I think further down, it tells us – and if you can confirm it from memory, we maybe don’t need to – it was three or four incidents which created a problem?

Elizabeth Evans-Jones: I don’t think there’s a specified amount of incidents that would create a problem. So you could actually have a problem – this is, again, the academic theory of service management. If – any incident that you do not have a resolution for or a workaround that would restore service could trigger the raising of a problem, and then that problem then should be investigated as to what the root cause is and then that root cause should be removed from the infrastructure through the change management process.

Ms Page: But in this document – and perhaps we can scroll down to see if we can find it – I think it’s right to say it was, in fact, three or four incidents which were defined as becoming a problem?

Elizabeth Evans-Jones: I don’t know, I can’t see that in the document. The theory is that any one incident can generate a problem and perhaps in this document it stipulates two or three. I feel – I don’t see on here where it says that.

Ms Page: All right, then let’s just stick with one or more incidents, then. The incident, we saw earlier, was defined as “any event which is not part of the standard operation of a service and which causes or may cause an interruption to or a reduction in”, yes?

Elizabeth Evans-Jones: Yes.

Ms Page: All right. Can I just have look at how that translates into – falls into the system. If we could bring up, please, POL00073280. If we go to page 5. Now, page 5 shows us what seems to be a typical record of a call in to the Helpdesk; is that right?

Elizabeth Evans-Jones: Yes, that’s what it appears to be, yes.

Ms Page: Presumably this is the sort of output of the PowerHelp tool; is that right?

Elizabeth Evans-Jones: Yeah, this is from PowerHelp, correct.

Ms Page: In this particular incident, we see that it’s a call in on 28 January ‘04, we see that in the middle of the top.

Elizabeth Evans-Jones: Mm-hm.

Ms Page: We can see there’s a box called “Problem Text” about halfway down, a little bit below halfway down, and this is a summary of what the caller says:

“Caller states that discrepancies are going through on the system. And this has been the case for 3 weeks in a row.”

Then it gives the amounts for the discrepancies. Then we also see a little below that, two lines below that, a text after the call has been closed and this appears to be a sort of summary of why the call is closed.

Elizabeth Evans-Jones: Mm-hm.

Ms Page: It says:

“Call Close by Diane Meah: NBSC issue, transferred for investigation.”

So that presumably is a typical closure if the caller has been referred to the NBSC?

Elizabeth Evans-Jones: That’s my understanding from the text that’s written on here, yes.

Ms Page: Is it right also that we would – we then can see below that, “Non-horizon Business” as the product and the description, and presumably that feeds in, again, then, to your metrics, does it, for how the call has been resolved?

Elizabeth Evans-Jones: Yeah, if you refer back to the table with the graphs, one of those blocks would be related to – calls classified as non-core or referred to – I don’t know what the terminology is, whether they’re referred to NBSC or whether it’s non-Horizon business. So it would fall in one of those blocks on that graph that we saw.

Ms Page: I think you told us, didn’t you, that, from memory, you didn’t particularly remember discrepancies being their own type of resolution?

Elizabeth Evans-Jones: Yes. No.

Ms Page: But we see here an example of how a call about discrepancies, is resolved as non-Horizon business, yes?

Elizabeth Evans-Jones: Yes, that’s what this is showing.

Ms Page: We don’t necessarily need to go to them but there are then, following this call, from the same office, which is Marine Drive, it’s a particular office which obviously this Inquiry is going to hear a little about, there are then a number of calls about discrepancies, which are all basically resolved by being referred to the NBSC.

So that is an example, is it not, of how calls about discrepancies would never turn into or, in this case, don’t appear to have ever turned into “incidents” or “problems”?

Elizabeth Evans-Jones: In this particular incident, yes. This wouldn’t have been investigated by Fujitsu. However, the comment on the bottom of the screen that I can see there is that the NBSC would then be able to refer that back to Fujitsu following their investigation. If you recall the incident flow, one of the inputs at the top would be from users or from the NBSC. So this could have been referred back into Fujitsu through – and I don’t know if it did but this could have been referred back through to Fujitsu from Post Office Account through the processes and the engagement that we –

Ms Page: Yes, I understand. It’s right. We can indeed see that there is a bit of back and forth between NBSC and the Horizon Helpdesk but, absent it being escalated from the Horizon Helpdesk, it can’t become an incident or a problem?

Elizabeth Evans-Jones: Or have been escalated through NBSC through Post Office into Fujitsu, yes. That’s –

Ms Page: Yes, I see. So NBSC could escalate it straight up the line, could they?

Elizabeth Evans-Jones: Yes.

Ms Page: All right. Thank you, those are my questions.

Elizabeth Evans-Jones: You’re welcome.

Mr Stevens: Sir, I think that’s all of the questions from the Core Participants.

Sir Wyn Williams: Well, thank you very much for coming to give evidence at the Inquiry and for providing a written witness statement. I understand you may have travelled from mainland Europe to give your evidence.

The Witness: I did indeed, yes.

Sir Wyn Williams: Yes, if that’s been inconvenient for you, I’m sorry. But I hope you’ll combine it with something which gives you some pleasure. So thank you very much.

The Witness: Thank you very much, appreciate that.

Mr Stevens: Thank you, sir. If we may have a 10-minute break for the next witness.

Sir Wyn Williams: Yes, certainly. What is that, sorry?

Mr Stevens: 12.10, if we may, sir.

Sir Wyn Williams: Yes, fine.

Mr Stevens: Thank you.

(11.57 am)

(A short break)

(12.10 pm)

Ms Kennedy: Good afternoon, Chair.

Sir Wyn Williams: Good afternoon.

Ms Kennedy: May I call Mr Amandeep Singh, please.

Amandeep Singh

AMANDEEP SINGH (affirmed).

Questioned by Ms Kennedy

Ms Kennedy: Mr Singh, as you know, my name is Ruth Kennedy and I ask questions on behalf of the Inquiry. Could you confirm your full name, please?

Amandeep Singh: My name is Amandeep Singh.

Ms Kennedy: You’ve given a witness statement to the Inquiry. If we could turn that up it’s WITN06660100. Have you got that witness statement in front of you?

Amandeep Singh: I do, yeah.

Ms Kennedy: If you turn to page 3. Is that your signature there?

Amandeep Singh: Yes, it is.

Ms Kennedy: It should be dated 13 January 2023; is that right?

Amandeep Singh: That’s right.

Ms Kennedy: Have you read through this statement recently?

Amandeep Singh: Yes, I have, yes.

Ms Kennedy: Is it true to the best of your knowledge and belief?

Amandeep Singh: It is, yes.

Ms Kennedy: If we could turn to paragraph 1 of that statement, so scrolling down. You said you worked on the Horizon Helpdesk support desk at Wakefield between September October 2000 and September 2001; is that right?

Amandeep Singh: That’s correct.

Ms Kennedy: What was your background prior to getting that job?

Amandeep Singh: So the background to the – me getting the job was it was my industrial year from university. I was studying computing at Huddersfield University, and we had to obtain a graduate work placement year, so the university found a placement for me. I wanted to do something that was a bit more hands on than what they initially found, so I found a role with ICL which was going to be supporting Epson Printers and I chose to take that role.

Ms Kennedy: So was this your first job?

Amandeep Singh: This was my first ever full-time role, if you like, yeah. I had worked part-time prior.

Ms Kennedy: When you joined, how many people were part of the Horizon Helpdesk support desk?

Amandeep Singh: Sorry, just to backtrack, I joined the Epson Helpdesk initially and, at some point during the year, it merged to become that – I think ICL and Fujitsu had some type of merger, and then it became – we were all transferred to the Horizon Helpdesk. Roughly, I think, there was maybe six to eight teams and each team had about maybe 12 – 10 to 12 members.

Ms Kennedy: Are those the squads that you were referring to in paragraph 2?

Amandeep Singh: Squads, yeah.

Ms Kennedy: How many squads did you say there were?

Amandeep Singh: I think it may be six to eight. I’m trying to really rack my brains. Between about six to eight.

Ms Kennedy: What did your role involve when you joined the Epson support desk?

Amandeep Singh: So my role was initially as a first line support engineer. We would support all Epson printer products that weren’t related – Mac related, if you like.

Ms Kennedy: How did that change when it move to the Horizon Helpdesk?

Amandeep Singh: So the role initially was supporting maybe technical people, and – so you’d get people in from organisations, people calling in or even just generally IT savvy individuals, if you like, that had issues with their printer, and we would just talk them through it. Sometimes there would be drivers issues or printer driver – we’d navigate them through software, how they’d install drivers. If we couldn’t then resolve their issues we’d then pass them on to a second line team and they would – again, they would – kind of more specialist and a bit more – maybe more technically able than what we were in the first line team. I’m sorry, the question of how it –

Ms Kennedy: What training did you receive when you moved over to the Horizon Helpdesk?

Amandeep Singh: So we were all told we were going to be moving to the Horizon Helpdesk. Initially, it was something that we weren’t aware of what we would be doing. But we got, I think it was on a few days training, we got to see the equipment, run some dummy transactions. We were in a room where we learned how to use the software. We were given a booklet on the transactions, how to navigate and to do things like selling a stamp, for instance, or it was routine transactions that maybe a postmaster would do. And that was the level of training that we received. Roughly off the top of my head, I think, I got two or three days training.

Ms Kennedy: You say in your statement, if we turn to paragraph 4, so over the page, that you think it was insufficient. You say:

“The support staff faced the initial challenge of basic training that was insufficient to fully support postmasters in the full array of tasks that Horizon was set up for.”

Amandeep Singh: Yes, I think initially it was useful because we hadn’t seen the software. When you’re on a phone call, you have to visualise what the postmaster is visualising, and what the – the transaction that they’re trying to do. But we were just given routine transactions. I think we did one time where we had to do the reconciliation task. I think we had – off memory, I think we did it once. But generally, it was how we would go about doing certain transactions and that was it. But we didn’t really know what the calls would be until we got on the call because this is really the real inception of the Helpdesk itself.

So until the calls started coming through we didn’t really know what level of support we would be providing postmasters, and the postmasters themselves quickly picked up how to do the transactions. It wasn’t something they were going to struggle with but that’s the level of support that we would get, I think. Where it was insufficient was it was the more complex transactions. I think they had foreign currency exchanges and how they put cheques through a system, and there was things that we didn’t come across originally.

So, again, it was learning on the job, and a lot of it was just trying to remember, look at your booklets and guiding a customer – guiding the postmaster through certain transactions, and it was generally at that level.

That’s what I meant by insufficient. It wasn’t more than just routine transactions, which is what the Helpdesk training was really all about.

Ms Kennedy: How many calls were you fielding from subpostmasters?

Amandeep Singh: It felt, initially at the start, we were sort of inundated really, to be fair. So there were – just a constant stream of calls tend to come. They did used to obviously peak on Wednesdays but it was a steady flow of calls constantly.

Ms Kennedy: Why did they peak on Wednesday?

Amandeep Singh: That was their reconciliation day, so that’s when postmasters would then do their – if you like, they’d balance the books, as such.

Ms Kennedy: Can you describe what that day was like, from your perspective?

Amandeep Singh: We would generally come in later because – we had different shifts, to be fair. So there’d be the normal shifts that were covering – they’d finish around 5.00 but you’d always have certain teams that would have to stay longer because we kept the desk, as far as I can remember, open for longer that day because you knew it was going to be a really heavy, heavy day. You could be on a call with a postmaster for a few hours trying to help them to reconcile, and that was very stressful days.

Ms Kennedy: You say at paragraph 5 of your statement you’d gone from dealing with IT savvy people, essentially, to people who had never owned a machine before and weren’t computer literate; is that right?

Amandeep Singh: Yeah, that’s right. I mean, like I said, this was early 2000s. A lot of postmasters had worked in their – you know, in their branches for decades, in some cases, and they’d never been around even a personal computer. So it was not only introducing, you know, personal computing into their lives, really, in some cases for the first time, but it was then also giving them exposure to “Here, now use this software”. And some of them wouldn’t – you’d have to explain to them what the mouse was, in some cases.

I mean they wouldn’t know what you meant by a mouse. I mean, it’s literally this is the age, the time we were dealing with – with certain people. Not everybody, but a lot of postmasters were elderly. Some of them – I mean, a lot of them weren’t IT literate at all. I mean, generally the public, you could say at that time, not many of them were either.

But – and that brought its anxiety and stress to postmasters themselves and as well as to us, because we had to explain sometimes maybe a complex transaction, but knowing that they themselves were not very literate in terms of just orientating themselves around the screen, trying to pick the right transactions, the right icons, and you’d have to describe the icon on the screen in detail. Go “Yes, press that. Now you’re seeing another screen”. And so it was really – trying to guide them as much as you could, and that itself did bring a level of stress because you were constantly aware that other calls were coming through, there was a backlog of calls, and you knew that that the call started somewhere but it would end somewhere. And it was trying to get them to that resolution point and sometimes you would just feel a bit deflated in terms of how can I get this person to that endpoint when they’re not really capable of sometimes getting there?

Ms Kennedy: What training, if any, did you receive in how to deal with people of different computer literacies?

Amandeep Singh: Nothing.

Ms Kennedy: Were there any particular types of problems that you were asked to be ready for, or examples of issues that subpostmasters may face that you were trained on?

Amandeep Singh: No, not that I can recall.

Ms Kennedy: At paragraph 6 of your statement, you say that the floor was quite a toxic place. Could you tell us a bit about more about what you mean by that?

Amandeep Singh: Just to elaborate on the point that I’ve made about the postmasters themselves. So, for me, this was my very first IT role so I was – dealing with people that were ringing in and wanting the help on their printers were generally people that tended to know about – you know, at the time it was Windows 95 or Windows 98 and you’d “You can install a driver?” “Have you checked this?” You know, how to run a clean cycle on a printer. All of this terminology, in many cases, was just over the head of a lot of post – you know …

And for me it was a learning curve because it was, without being too crude, it was a job and I thought “Well, I’m a uni student, I’m going to go back to uni, I’m just going to see this out and see this is what the world like”. But it was toxic because the other members of staff that were, if you like, the second line team, the whole second line team had been abandoned and everybody had just merged into this Horizon Helpdesk.

So there was a hierarchy of the second line team – so we would never interact with, as first line engineers on the Epson team, because we really, some of them, we would hate having to put calls through to them because they would almost belittle you in some way when you would pass a call to them. Like “Can you not deal with this yourself?” There was a hierarchy of individuals, to say we are really talented in engineers. They hated us passing calls through to them.

So there was that dynamic where you didn’t really associate with those guys well. And then when they were all brought in, everybody was equalised and on the same level, that caused a great deal of animosity. And then layer on top of that, you’re now not supporting maybe graphic design agencies or media companies as second line engineers were doing, and now you’re supporting, you know, an old lady in Wales that doesn’t know what even a personal computer is.

It felt, I think a lot of them felt like the role was beneath them, and that animosity, that toxicity, it just grew and grew. And people were – it became a bit like people were almost on the calls and they were almost shouting about “Oh my God, I’ve got this person on a call”, and this – and it became almost comical to watch people frustrated and throwing their arms about and making a scene about supporting somebody who can’t do. And obviously they weren’t projecting it to the customer because it were going on mute, throwing their arms up, “Oh look at this person, I can’t believe I have got this, they don’t know this, they don’t know that”, and you’d get had a lot from the members of staff that were, you know, the second line engineers, if you like.

And it just created an atmosphere that were just – you didn’t really want to be there. The people and the teams didn’t want to be there, and, like I said, it were just going through – for me, just going through the motions of getting through each day.

Ms Kennedy: If we turn over the page, still in paragraph 6, you mentioned some of the comments that you heard while you were there. Do you want to tell us a bit about that?

Amandeep Singh: Yeah, because I wanted – the reason why I got in contact with the Inquiry itself was because it felt a little bit like – I don’t know how much that the – it was going to be an inquiry about senior management or maybe people looking away or people from top-down dictating practices or something, and I wanted to just give you my real-world experience of what it was like just on the Helpdesk on the floor.

It wasn’t like this Big Brother element of senior management, just my opinions of day-to-day of what it was like. And what I wanted to get across was you had that element of where you had the teams merging into one, supporting the postmasters, that resentment towards the role, that they’d been now forced to take on. And then you had another layer upon that, with the cultural issues in some cases. Many of these people were supporting – Asian postmasters, not to put it in any blunter terms than that, but sometimes an Asian postmaster they would ring up and they’d be it like “I’ve got a £2,000 discrepancy, I’ve got a £5,000 discrepancy”, in some cases you’d get wild figures like 50,000 or 100,000 and sometimes these figures, people were quoting were more than most of the salaries, annual salaries of the individuals that they were supporting.

And if you like, so if you layer upon that, they’re doing a job they don’t want to do, then someone rings up and says £15,000 is missing from my account, and people were like “Oh, here’s another one calling”.

And you’d get so many of these calls, not just from Asian subpostmasters but from everywhere, all walks from the UK, people would be calling in and they’d be saying these figures and it’s more than, you know, a monthly salary, more than an annual salary of a lot of individuals and they’d think “Where’s this money gone?” And it just build that mistrust.

And I think at the time, there was always stories in the newspapers of somebody maybe frauding or defrauding or doing something. And if somebody rings you up and says £50,000 or £10,000 is missing, and you’d be like – even – you know, you’d question it and say, “How can you miscalculate 10,000 or 2,000?” It’s not a couple of stamps here or there. You can’t reconcile to £20 or £30. But these were huge figures that people were quoting.

But – and I think that’s what I meant. Once that story got about, once somebody said, “Oh, I’ve got another Patel”, and then you could just never get away from that whole “Patel” thing. And it would be like who could come up with the most outlandish story, “Oh, I’ve got a Patel, got another Patel”, you’d just hear it constantly on the floor and me obviously being from an Asian background, there was me and another gentleman called Zubair we were the only two ethnic minorities on the support desk at all, on the whole entire floor. There was a Chinese gentleman or someone from a Chinese background, Peter, and they were the only people of sort of colour on that floor.

And at no point was anybody reeling it back and saying, “What is this – you know, the language that was being used?” And my grievance with it was that it was a case of why don’t we just focus on the individual or the actual – this cultural dynamic, this mistrust, that was just feeding through. It felt like some individuals could never get beyond that, could never look beyond that and try and do the role they were instructed to do.

And this is one of the prime reasons that I wanted to, you know, get in touch with the Helpdesk itself.

Ms Kennedy: How did that make you feel, working in this environment?

Amandeep Singh: I told myself every day that I’m here just for the year and I’m just seeing it, and I was paid almost twice as much as nearly all the other graduates that I knew, so I knew that I was well paid. And, for me, I thought “It’s okay”, and speaking with – from Indian parents, to be fair, my parents, my mum and dad go “This is just work, this is what it’s like in the outside world”, you would just get told – and that was really depressing, to be fair, to be told that. And I just thought “Wow, I’ve got a whole lifetime of this ahead me and this is what it’s going to be like so I’d better just get used to it”. And I’d just go into work and be just like okay.

And nobody ever said anything racial to me, I would just fit in with the team. I was with the most vocal team, which was Squad A. But nobody ever said anything personal to me, I fit in and I could hold my own, I’m quite thick skinned, I grew up in that environment, so it wasn’t difficult for me in many ways in such as I look back now, as somebody who has been and worked in the industry for almost 20 years and look back and think – you know, now having two boys and having young kids thinking how difficult I would feel if they were in that environment.

But me personally, I think I just find it harder now to look back than I did then, whereas – when it was just a case of: let’s just get through the day. It’s another day. I’m earning good money. Let’s just move on. And that’s what it was. But it was difficult.

I did know that a lot of conversations were going on, and it was a case of when I would walk into a room sometimes it would go quiet, and I knew some things wanted to be said by certain individuals. So I would almost make an excuse to leave to let them complete their conversation and then go back, because I know that they wanted to say something. And it was a case of managing that environment, for me. But I really took it as a point of: this is something that I’ve got to go through and learn. And that was really sort of how I navigated my days, really.

Ms Kennedy: Turning back to your statement, if we look at paragraph 7, you say:

“As for their reconciling issues [the subpostmasters], when we could not help them with their accounts, this would mean we spent a few hours on the phone going through each transaction and trying to figure out where the financial discrepancy was. We would eventually give up, and we were advised to write off the loss as a ‘discrepancy’. This was a word you could hear from every agent’s calls.”

Do you want to elaborate a bit more on your experience with this?

Amandeep Singh: Yes, and just to caveat what I’m saying, I did feel that every agent, no matter what they said, they did try their best to try to get – to try to help every postmaster that, you know, they called up. But the Wednesdays days, you would – and bearing in mind that none of us were from an accountancy background, we were just IT people, but we would almost be bookkeeping live with an individual for an entire week’s transactions trying to get down where did they get this discrepancy from. So it would be if you like how many stamps did they sell? How many foreign currency transactions? So these are the transactions, that’s what you’re supposed to have. We would go line by line through every single transaction trying to understand where did this discrepancy come from. That’s why the call would take hours. This is why you had to almost, you know, physically build yourself up sometimes for calls when you knew, when somebody would call and go they’ve got a discrepancy for a few thousand. And you know right, okay, this is going to be a difficult one. And you’d go through all these transactions.

And the postmasters themselves were always quite frantic. They were, you know, they were so stressed. “How do I get this? How have I got this figure? How am I going to reconcile this account?”

And so, you know, we would work with them for hours. We would really try our best to get down to it. And then, you know, we couldn’t resolve it. We’d go to – we’d have a team leader, sometimes floor walkers, and ask their opinions, “Have you checked this? Have you checked that?” We’d go back and try and resolve it, and if we couldn’t, it would be like okay, it’s a discrepancy. Write it off as a discrepancy. We can’t really do anything more beyond that. And it just almost became the norm, in a way. And you’d have postmasters did say, “I had one last week, I was like” – someone had £46, it was small figures. They’d be like, “I will put money in myself, just to circle it. Just to square the circle, if you like, just to get it to a zero balance.” They’d be like, “I’ve been doing it now for weeks.”

And it was only when they got these extreme figures, these big figures, that they would call in. And then you’d find that that’s when they’d need help. Sometimes when they were small figures they’d tell you, “Oh we’ve been putting money in ourselves just to get it to zero.”

Like I said, you could just hear the word “discrepancy”. It was probably the most used word, as well, on every call. “Oh, have you got a discrepancy?” Like I said, it quickly went from “How do I do this transaction?” After a few months, people knew all the transactions. “How do I reconcile them?” Then nearly everything was just discrepancy, discrepancy, discrepancy. That’s what the calls were really about. People just not being able to, you know, reconcile their accounts to zero.

Ms Kennedy: When you say, “We were advised to write off the loss as a discrepancy”, who was advising you?

Amandeep Singh: So now, I really tried to rack my brains on this one because we – because there was a – we had a management team that were in the helpdesk. They were in the sort of – the way that the helpdesk was located, you had the managers that would sit in the middle of the helpdesk. And I was trying to rack my brains and think who was telling me? And I remember it was – sometimes it would be, like, just one of the people in our team that were the most able on the software. And you’d cross-reference it with some of your colleagues. And then I think they’d put in some team leaders type in place, because the managers themselves, they didn’t know nothing. They never touched the software. They didn’t do any training. You’d only go to your managers if you wanted to get a holiday. You’d go to them and go, “Can I have a holiday booked?”

And then managers were acutely aware that the floor was struggling, so they almost strategically picked out people out of each squad that were the most able on the software, and sort of made them like floor walkers or team leaders or advocates, if you like. And you’d go to them and say: “Right, okay, I’m struggling.”

And they would go, “Well, there’s nothing we can do. This is a discrepancy.”

And that was it. It was never the managers. I mean, like I say, other than signing holidays, I don’t know what they did. They weren’t – like these days, if you look at call centres, you have people listening in to calls, reviewing calls. In the year that I was there, I honestly can’t tell you what they did other than sign our holidays – you know, sign off holidays, or we’d ask them for that. And there were a good few of them. There were a good five of them – five, six, seven of them.

And it was a gripe that most of the engineers had as well: that what is their role? What do they do? Because we needed help, we needed guidance, and we didn’t really get it. So it was your colleagues. To answer your question, it was kind of your colleagues, and then the floor walkers which were normally – which were, again, your colleagues that you’d go to. So there wasn’t anyone in a senior or a management role that you’d told. And that was again one of the reasons why I wanted to contact the – because rather than saying it was some sort of mythical big bad manager who was telling you whatever, or guiding you, in the very initial year that I was there when it was set up, it felt very much like a rudderless ship, really, and you were just guided somehow on your own.

And I think – it probably stems from all the other issues I raised – just the lack of management in that interaction across the floor.

Ms Kennedy: Do you think genuine issues with Horizon were missed because of the toxic culture?

Amandeep Singh: Being there only a year, it’s very difficult to ask that question. To answer that question, sorry. I think it didn’t help, it really didn’t help, because if people were genuine having software issues, but if you’ve already got a pre-built prejudice that you can’t trust the people or the people are incompetent – and that’s really, like I said, the nub of the issue for me – is if you’ve already made a judgement call on the people that you’re supporting as incompetent or corrupt in some way, it would take a lot of people to go, you know, that the software has an issue.

Because I think people were – we were much happier, on the floor, to push down on to the postmasters and go “This is your issue”, or “You’re not correct”, or “You’ve got the issues”, than anyone on the floor going – pushing this upwards and going, “Is this is an issue here? How can we have so many of these calls?” Like I said, we didn’t know who to push up to. And the management were just not visible. Like I said, I just don’t know what they did.

So you can see it must have taken almost like a snowball effect on – for someone just to look into this issue to go: “Surely we can’t have this many discrepancies?” So you can see how it must have just snowballed. And like I said, I was only there for a year, and maybe it just grew and grew, and then eventually, you know, just through the number of issues and discrepancies, that’s how it must have got through. But I don’t think the people’s, you know, pre-built prejudices, I don’t think they helped at all because they could never empathise with the postmasters. They could never understand their issues. Even when they were upset and crying on the phones – which we had all the time. Really upset individuals trying to understand, trying to, you know, get their accounts to zero. But I think it’s difficult. It’s very simplistic to just say that. I think there was other issues involved, but I don’t think it helped.

Ms Kennedy: Is there anything else you wanted to say about your experiences to the Inquiry?

Amandeep Singh: Um, like I said, they would be more personal on me in terms of how I felt as an individual, and I don’t want it to blur the fact that this is an issue about postmasters and basically about the Post Office itself, or – and, you know, their own issues as opposed to how I really felt. I think just – to me, it was a lot of issues at the time, just a lot of issues of just, like, the postmasters, their technical capabilities, a lack of absence of management, of people. And, you know, I know that now they’ve got like first, second line teams and problem management and change management, and all this. When I was there, they had just that one team. It was immature. Now I can look back after 20 years in the IT industry, I can look back and see just the people themselves, the managers were not qualified to do the job. The individuals, some of them, should never have been supportive.

It looks just, you know, with hindsight – and it’s easy with hindsight to say these things – but there were just culpability on so many levels, on so many areas. And it’s probably good to have this, the review, to look, just from an organisational point of view, of how these structures, when they’re not there, that’s how issues like these can arise. When you don’t have the change management, when you don’t have the problem management, when you don’t have analysis. And we didn’t have much of that stuff going on, and we had an absence of management, of people not probing in to look as to why there are issues.

You know, this is why, you know, since then, you have this ITIL standards or the Service Desk. You just have things that didn’t exist at the time.

Ms Kennedy: Thank you, Mr Singh. I don’t have any more questions for you. Mr Stein has a question.

Questioned by Mr Stein

Mr Stein: Mr Singh, I represent a large number of postmasters and mistresses. Can we just describe, please, the area where you worked so that the Inquiry can grasp whether you worked in booths or whether you worked in a large, open-floor space. Could you just describe the area you worked?

Amandeep Singh: It was the large – a large area. We had our own desks but they were segregated slightly by the squads that you worked in. So you sat with your squad, but it was very much an open desk. There weren’t private booths. So it was very much open.

Mr Stein: So if somebody is shouting out from your squad, would another squad be able to hear that, across the floor?

Amandeep Singh: Yes.

Mr Stein: You mentioned the floor walker system, and you’ve also spoken about managers. In your statement, you say this: that when people were being vocal and toxic in what they’re saying, that they were unchallenged by managers. And I quote from your statement that they looked away the managers looked away, “as all Asians were called Patels, regardless of surname.”

Amandeep Singh: Mm-hm.

Mr Stein: Did you ever see a manager discuss the behaviour of the individuals who spoke in that way?

Amandeep Singh: No. And I now, being obviously 44 years old and not someone who was like 21 at the time, I would challenge that. But then I think these days it’s very rare you would need to challenge that sort of behaviour. But at the time, there were some very strong-willed characters there who had almost, you know, roughshod – they could ride roughshod over whatever they wanted to say. But it was a case: “Oh, he’s a character. Oh, they’re characters.” And we just let them go. We would just leave them to it, if you like.

They never challenged them but then, like I said, other than signing our, you know, holidays – asking for holiday, I don’t think they did anything, those managers. I look back now and think, you know, the bureaucracy that must have existed to have so many managers which then led on to other managers, and then neither of them, you couldn’t really define what their role actually was.

Mr Stein: Two last points. Do you by any chance remember the name of the manager that you were directly dealing with? Was there a single individual that you would have described as being your manager? And if so, can you remember their name?

Amandeep Singh: I can remember their name. So mine would have been Geraldine McEwan, I think it would have been.

Mr Stein: Thank you. Lastly, then, you’ve said that you don’t want to confuse issues between the effect on you versus what was happening to the subpostmasters and mistresses. Was the effect upon you, what you were going through in that period of time, was that something that inhibited or stopped you from, as an example, trying to take it any further within the organisation?

Amandeep Singh: Um, taking it further was never a thought in my mind. I’m going to be bluntly honest with you. I couldn’t wait to get out of that role quick enough, and I did see it as the fact that this is a rite of passage for me. It’s something like coming of age, doing my role, doing my time. And the best way to describe it is like if you’re in prison and you’ve got the tally charts and you are crossing the days off to go: right, I’m going to leave on that day. And like I said, I was well paid. And I didn’t want people to think that it affected me, the language, as well, on the floor. So I didn’t want anyone to think that I’m just weak, in some ways, or that I’ve got an issue with it, or I’ve got a chip on my shoulder, or this language, or – it was easier for me to just go: let’s just see it through. It’s fine.

One thing I do want to mention is that it’s very much an issue of I feel, having worked at, like I say, predominantly within the south, within London, within the banking industry, that it’s, to me, it was very much a cultural issue of Yorkshire, Wakefield, of communities that don’t mix and are mistrusting of each other.

And this is why I wanted to raise the issue of this why people that were hiring, the incompetence level of it. If you’re supporting people from Wales and villages in Wales and in Scotland, and there predominantly is a lot of Asian people owning post offices, is for you to understand the people, as well, that you’re going to be supporting. And, you know, be able to put yourself in their shoes or walk, you know, in effect walk in their shoes and understand their life situations. And the people that they were, having supported them, could never do that, and are almost incapable of doing that. And I think that’s one of the issues that I wanted to sort of raise as well.

Mr Stein: Thank you, Mr Singh.

The Witness: Thank you.

Ms Kennedy: Chair, I don’t think there are any further questions from any further Core Participants.

Sir Wyn Williams: All right.

Well, Mr Singh, I’m very grateful to you for drawing these matters to my attention and for making contact with the Inquiry, and being determined to give oral evidence about these things. So thank you very much.

The Witness: Thank you.

Ms Kennedy: Chair, that concludes the witness evidence for today. We’re back tomorrow with Mr Andrew Dunks.

Sir Wyn Williams: Yes, fine. All right, then. 10.00 tomorrow.

Ms Kennedy: Thank you.

(12.46 pm)

(The hearing adjourned until 10.00 am the following day)