WITNO6060100
WITNO6060100
Witness Name: Graeme Seedall
Statement No.: WITNO6060100
Dated: 07-03 -2023
POST OFFICE HORIZON IT INQUIRY
FIRST WITNESS STATEMENT OF GRAEME SEEDALL
I, Graeme Seedall, will say as follows:
This witness statement is made to assist the Post Office Horizon IT Inquiry (the
“Inquiry”) with the matters set out in the Rule 9 Request dated 14 October 2022
and a list of follow up questions sent on 19 December 2022.
1. As part of the Rule 9 request, I was sent a set of questions and several
documents to consider as reference material for those questions. I answered
the questions to the best of my recollection, using the documents to assist me
where possible. A subsequent set of questions was later sent to me which
related to my answers but requesting clarification and further detail. These
questions were accompanied by a further set of documents for reference
which in many cases revealed a deeper understanding and greater recall to
me. The two sets of questions have subsequently been amalgamated with
linear numbering in order to create a narrative. For this reason there may be
1 of 37
WITNO6060100
WITN06060100
multiple statements relating to events that are slightly contradictory or at very
different levels of detail, depending on my level of recall at the two separate
times of questioning. Moreover, given that the questions relate to incidents
which took place two decades (or longer) ago my recollection is limited,
particularly at the detailed level where my recall is limited to the reference
material I have been given. I have identified where this is the case.
Background
2. I joined the Post Office in 1980 and was a Postal Officer on the counter at
Boscombe Branch Office until moving to the first automation pilot (Counters
Automation) in 1989. I dealt with user issues on the Post Office Help Desk and
continued in that role when Counters Automation Thames Valley pilot was
replaced by the ECCO+ system (A branch office accounting system). I moved
to the Automation Development Team in 1995 and was drafted onto the
Benefits Agency/POCL joint working programme to participate in the business
side of EPOSS development. After moving onto various subsequent
automation projects, I left POCL in 2009 following some personal and
professional issues.
3. I have been asked about the extent of my training and qualifications prior to
joining the Post Office. I joined as a postman, when I transferred to the counter
I undertook their internal exams and training course.
2 of 37
4.
WITNO6060100
WITN06060100
I have been asked how I came to be recruited in the field of automation in
1989. I applied for a job through the internal recruitment process, although the
job related to an automation programme my role was related to business
operations. I am not technical and have never worked in a directly technical
role, such as systems development or technical testing, for the post office.
. I have been asked what technical knowledge or expertise I had at this time. I
had no purely technical knowledge. I had a good working knowledge of the
system and the front office processes.
. I have been asked to describe the nature of my role on the Post Office Help
Desk during the Counters Automation Thames Valley pilot. The Post Office
Help Desk was the first point of contact for issues arising from the pilot offices
which apparently related to automation, subsequent resolution of issues either
through direct advice or via the technical helpdesk, which was at that point
internal to the Post Office (Post Office IT section).
EPOSS
7.
I have been asked to describe the purpose and function of EPOSS. The
development of EPOSS function was required in order to support Branch
Offices, which had already had accounting functionality automated via ECCO+.
I had operational knowledge of that system from working on the Post Office
Help Desk and I was at that time in the appropriate team to provide resource
3 of 37
WITNO6060100
WITN06060100
into the newly formed BA/POCL programme. Additionally, I had been
requested as a specific resource for a RAD approach to developing EPOSS.
Initially I worked alongside an ICL Pathway contractor, who was a previous
colleague from the Post Office Help Desk, within the ICL work domain. This
approach was subsequently dropped by ICL and I moved into a joint BA/POCL
business assurance team working to a manager.
8. The purpose of EPOSS was to replace the functionality of ECCO+ and
therefore provide continuity to the Branch Offices which had already moved to
automated accounting practices using ECCO+. The functional scope was
therefore meant to be equivalent to ECCO+
9. I have been asked what I understood the RAD approach to entail. Until it
started I did not know. Once it began I was placed within the technical
development team and my role was to provide operational input to and
feedback on development of the initial user interface and attendant processes
for the new EPOSS system.
10.1 have been asked why this approach had been adopted for the development
of EPOSS and the nature of my involvement. It was a Pathway decision taken
without input from POCL.
11.1 have been asked when I was involved in the RAD approach to developing
EPOSS. I believe it must have been 1996.
4 of 37
WITNO6060100
WITN06060100
12.1 have been asked what (if any) concerns I had about this approach. None, it
seemed to be working well at the time.
13.1 have been asked to identify the ICL Pathway contractor with whom I had
previously worked on the Post Office Help Desk, and the nature of their
involvement and technical knowledge. His name was Richard Sloggett. He
was initially the lead developer of a prototype for EPOSS. I don’t know what
his technical knowledge was. He was later replaced by Steve Warwick,
another contractor who I also knew as he had worked on the technical service
desk for Post Office Group IT.
14.1 have been asked why the RAD approach was dropped. I don’t know why it
was dropped, it was a Pathway decision.
15. I have been asked what ICL Pathway was proposing to do in its place. They
brought in a new contractor and started to design and develop behind closed
doors, with the more formal review sessions as the system developed.
16. I have been asked to whom within POCL I reported when I was involved in the
RAD approach. I don’t clearly remember, but I think it was lan Gair who was
based within POCL or possibly John Meagher who was part of the joint
venture programme.
17.1 have been asked when I was transferred into the POCL assurance team. I
think this was probably late 1998, possibly early 1999.
5 of 37
WITNO6060100
WITN06060100
18.1 have been asked to identify my manager within POCL at this time. It was
Ruth Holleran.
49.1 have been asked how the branch office balancing and cash accounting
operated prior to the implementation of EPOSS. In Branch Offices balancing
and cash accounting were operated through the ECCO+ system. I am not
completely sure but I think that at the tine ECCO+ had been rolled out to all
Branch Offices (only) (i.e. those directly owned and operated by POCL).
20.1 have been asked if it was my understanding that, at the time, EPOSS was to
be installed only in Branch (or Crown) Post Offices. I don’t remember that but
it may have been the case for a while, at least for initial rofl out and trial
period.
2
—
.I have been asked what knowledge or experience I had, at that time, of the
accounting practices and procedures in post office branches owned and
operated by subpostmasters. I had none, other than supporting them during
my tenure on the business help desk for the first widespread automation
programme (the “Thames Valley" counters automation programme) around
1990 — 1992. I do not believe their accounting practices and procedures
differed significantly from those in Crown offices at that time.
22.1 have been asked to what extent the needs of subpostmasters were taken
into consideration during the development of EPOSS. I believe the intention
was to streamline any slight differences in the procedures into a single
common business process.
6 of 37
WITNO6060100
WITN06060100
23. I have been asked in which ways the office balancing and cash accounting
changed as a result of the implementation of EPOSS. I don’t remember at the
detailed level. The requirement was for functional equivalence and for the
need for retraining to be minimized. However, I think some functions were
subject to a redesign during EPOSS development. Additionally, accounting
functions were required that integrated with newly automated client
transactions, specifically APTs and benefit payments.
24.1 have been asked about what involvement I had in specifying the functional
requirements of EPOSS and what challenges or difficulties I experienced in
defining the relevant requirements. As I remember it, there were no detailed
functional requirements, so none.
25.I have subsequently been asked to explain what I meant by “the requirement
was for functional equivalence and for the need for retraining to be
minimized”. I believe that is what the requirements stated at the time. As I
remember it, the business requirements were extremely high level.
Presumably a copy of them still exists somewhere?
26. I have been asked what my role as Acceptance Test Manager for EPOSS
entailed. I don't remember the detail of this. I know I was involved with
reviewing the ICL Acceptance documents and for discussing the issues
arising from testing along with my then manager Ruth Holleran, ICL/Pathway
representatives and other interested parties such as representatives from the
central accounting team in Chesterfield. I was not involved directly in the
testing of EPOSS.
7 of 37
WITNO6060100
WITN06060100
27.1 have been asked what I recall of the issues which arose in relation to the
testing of EPOSS. There were a lot of technical issues arising from testing
which were regularly reviewed jointly by members from POCL and
ICL/Pathway.
28.1 have been asked what challenges I encountered when seeking to assure the
design of EPOSS. I don’t remember there being any design documents nor
any contractual acceptance criteria relating to design. As far as I remember
the approach that ICL/Pathway adopted was to create Acceptance Criteria
that related directly to (i.e. mirrored) business requirements.
29.1 have subsequently been asked what challenges the absence of design
documentation presented for the business assurance team. During
development this was a real problem but once the system was in testing users
were exposed to the system and raised any design issues that caused
operational problems.
30.1 have been asked to what extent I had an adequate understanding of how
ICL Pathway would meet the business requirements of POCL in relation to
EPOSS. As stated before, the requirements were so high level that this was
not seen as an issue.
31.1 have been asked to describe the results of Model office and End-to-End
testing of EPOSS conducted in autumn 1998. I have no memory of this in any
detail. I do remember there were a lot of outstanding issues and there were
known issues relating to lost transactions and cash account imbalances which
8 of 37
WITNO6060100
WITN06060100
appeared to have been created by the system or by the reference data that it
was drawing on.
32.1 have been asked what I understood to be the causes of cash account
imbalances identified during the testing of EPOSS in autumn 1998. Again I
have no detailed memory of this. I believe that at the time it was shown that
the system was capable of creating imbalances in the cash account through
hidden processing errors or from reference data issues.
33.1 have been asked what effect cash imbalances had upon the integrity of data
being processed by EPOSS. I’m not sure I understand this question. I believe
that it was possible for cash account imbalances, seemingly created by
internal processing errors or reference data issues, to be presented as if they
were an actual shortage or surplus in the account.
34.1 have been asked what steps I took in December 1998 to address these
imbalances. Again I am unsure of the question here. The occurrence of
imbalances was documented and discussed at many levels by all parties
concerned. I believe that a process was created for identification and
correction of these “rogue” transactions or value discrepancies that resulted in
unsubstantiated imbalances and that it resided in the service management
domains of both POL and ICL Pathway.
Acceptance of the Core System Release
35.1 have been asked to describe the nature of my involvement in the
Operational Live Trial of the Horizon System. I don’t remember clearly but I
9 of 37
WITNO6060100
WITN06060100
think it was limited to attending lower level management meetings, receiving
summary reports and helping my then manager Ruth Holleran with impact
assessments of incidents raised.
36.1 have been asked to explain what I understood about cash account
imbalances observed during the Operational Live Trial. I don’t remember the
detail but having read the documents I recall that the interface to TIP was a
particular problem as was initial office data migration at set up.
37.1 have been asked about the nature of my involvement in raising Acceptance
Incidents. I don’t remember actually being responsible for raising Acceptance
Incidents, just for assisting in assessment of their impact.
38.1 have been asked about my understanding of the nature, cause and severity
of Acceptance Incident 376 (Al 376). I do not remember the detail at this level.
However having read POL00030393, I do remember in general terms it being
discussed. The nature and cause were presumably as described in that
document and the severity, although individually low within the test
environment as it was fixable with some data manipulation, was immense on
a business wide scale - particularly if there were underlying technical issues
which had the potential to continue to cause data corruption issues in future
despite the fixes identified by ICL Pathway. This was the concern at the time.
39. I have been asked how Al 376 came to be identified. After reading the
document I believe that that particular version of data corruption was identified
by the end-to end tests performed (i.e by the TIP team).
10 of 37
WITNO6060100
WITN06060100
40. I have been asked about the business impact of Al 376. As described in
paragraph 38 above the potential business impact was huge.
41. I have been asked what remedial action I proposed. None as it was not within
my responsibility to do so.
42.1 have been asked for my assessment of ICL Pathway’s proposed
rectification plans. Along with other business representatives I felt that while
individual incidents were being corrected by ICL Pathway, similar incidents
causing data corruption may be possible in future if there were issues in the
underlying system and that this was quite possible given the number of
incidents raised and the nature of the responses by ICL Pathway.
43.1 have been asked about the nature of my involvement in the acceptance
workshops carried out between August to September 1998. I don’t remember
the detail but it was probably to attend lower level meetings and assist with
impact assessment of incidents raised and business assurance of remedies
provided.
44.1 have been asked what my role as a “business expert” in relation to Al 376
entailed. I was mainly present to support my then manager Ruth Holleran and
assist with impact assessment and business assurance of remedies provided.
My area of expertise was in the user functionality of EPOSS not the interface
between the ICL Pathway system and TIP. The identification of me as
business expert for that area in document POL00028342 is incorrect, an issue
which Ruth Holleran seems to have corrected by including Martin Box (from
I TIP) as well as herself (also with central accounting expertise).
11 of 37
WITNO6060100
WITN06060100
45.1 have subsequently been asked if not the “business expert” what was the
precise nature of my role in relation to Acceptance Incidents 376 and 378. I
worked for Ruth Holleran, who was the business “owner” for this strand of
assurance. As such, my role included supporting her in all areas, providing
administration and judgement on issues where relevant. Also assisting in
holding meetings with other business experts.
46.1 have been asked what I understood the role of Peter Copping to be. I do not
remember Peter Copping.
47.\ have been asked what steps I took to obtain clarification of the root causes
of the fault. Personally, none. This was a known problem with a technical
interface that was beyond my area of expertise. The solutions were jointly
worked through without my detailed involvement as far as I remember. I
notice form the documents relating to the Acceptance Resolution,
FUJ00118194 & FUJ00079178 that I am not included in the distribution list.
48.I have been asked to describe my role within the “Working Group” which was
assigned responsibility for reviewing the “TIP Incident Status Report” and
reporting progress and issues back to the joint workshop. As described in
paragraph 44 I was a support to Ruth Holleran and therefore placed into
several meetings which underpinned the assurance and acceptance process
but may not necessarily be my area of knowledge. This was one of those
areas which I was initially involved in but only to oversee that a resolution path
was being followed.
12 of 37
WITNO6060100
WITN06060100
49.1 have been asked what I understood ICL Pathway to be doing to obtain
clarification of the root cause(s) of the cash account imbalances. As I
remember it, they were supposed to analyse each incident and give an
explanation, which may or may not result in proposed changes to the
underlying code, system process or reference data.
50. I have been asked to describe the nature of the “fix” which ICL Pathway
proposed to ensure that the cash account did not “lose transactions”. I think at
that time their proposal was one of simple additional reconciliation checks. I
think the incident developed beyond this... by which I mean that I believe
further investigation and further similar incidents led to a fuller understanding
of the issue.
51. I have been asked to describe what I understood the 3-level data integrity
check proposed by ICL Pathway to entail. I don’t recall but this would probably
have been covered off by the POL technical representative as well as people
from the TIP team.
52.1] have subsequently been asked to identify the POCL representative to whom
\ refer. That would probably have been Jeremy Folkes, initially.
53.1 have been asked if the 3-level data integrity check addressed the underlying
cause(s) of the cash account imbalances. I don't think so, I think it was just a
way to check that the data streams matched.
54.1 have been asked what involvement I had in scrutinising the design of the
integrity check. I believe I had none.
13 of 37
WITNO6060100
WITN06060100
55.I have been asked about my role in the “interactive walkthroughs” of the high-
level design for the data integrity check developed by ICL Pathway. My role in
the interactive walkthroughs of the high-level design would have been to
attend the meeting on behalf of Ruth Holleran and to ensure the business
experts were satisfied with the solution being offered by Pathway. As
previously stated, the TIP interface was not within my area of knowledge.
56.1 have been asked to describe the closure criteria which were agreed in
relation to Al 376. I believe these changed as the issue developed.
57.1 have been asked what role or input I had in relation to the agreement of the
closure criteria descried at p. 9-10 of FUJ00079178. I would have had no
input to the closure criteria which would have been a proposal from Pathway.
My role would have been as described above, that is one of overseeing the
process and ensuring the business experts were satisfied with the proposal
coming from Pathway.
58.1 have been asked to describe my understanding of the closure criteria agreed
for Al 376 in late September 1999 and whether I considered these were
adequate. I assume that the closure criteria agreed for AI376 in late
September 1999 are those listed in document FUJ00079178. I don’t clearly
remember my perception at the time but I believe that all involved POCL
representatives were nervous that these errors could continue to be present
after the checking period was ended but that there was no option but to agree
a way forward given the contractual obligations of the acceptance process,
and that these criteria were the best we could achieve at that time.
14 of 37
WITNO6060100
WITN06060100
59.1 have been asked to describe the reconciliation controls and/ or processes
which were agreed between POCL and ICL Pathway as a condition for
downgrading Al 376. Again, I believe these changed later. I think that this was
probably before the relationship of the issue to reference data was realized.
Roil Out of the Core System Release
60.1 have been asked to describe my involvement in monitoring ICL Pathway's
compliance with the closure criteria for Al 376. As far as I can recall, I was
involved at the start of this process, attending fairly technical meetings and
reporting up to more senior managers. As time went on the various threads
were monitored and managed within the appropriate specialist domains e.g.
TIP team, Reference Data team, Contracts team etc. My role in assurance
changed around this time but I can’t remember when or at which point I was
no longer involved.
6
.At some point I was given a role of creating a process and stakeholder group
for future change management. Although I recall being involved at the start of
the management of the TIP interface issues I do not clearly remember the
detail at the end so it is possible that my involvement began to be reduced
during that time.
62. I have been asked what I understood to be the root cause(s) of continuing
cash account discrepancies in the autumn of 1999. From memory, and also
from the documents you have provided, there were several root causes, with
new ones still occurring.
15 of 37
WITNO6060100
WITN06060100
63.1 have been asked what concerns I had from a business assurance
perspective about the emergence of new root causes for cash account
imbalances. As described above in paragraph 58, I and my POCL colleagues,
such as Ruth Holleran and Martin Box, were very concerned that new root
causes for cash account imbalances were still emerging during the
acceptance process.
64.1 have been asked to describe the issues which emerged in relation to the
quality of reference data in November 1999. At that time POL did not have a
functioning process for providing robust reference data and had not been
aware of the serious consequences of that on the system processes. POL
had, to some extent, been in denial of the responsibility to provide a suitable
feed of reference data to ICL and ICL had developed processes that had not
been suitably tested against the huge number of business rules and changes
to those rules. The issues that arose when the relatively unmanaged and raw
business data met those independently developed processes were serious
incidents in accounting data integrity (amongst others, for example automated
processes failing).
65.1 don’t clearly remember my role in addressing issues relating to reference
data but I think that I was involved at the start of the issue, attending meetings
and supporting Ruth Holleran when required, but dropping out of the detail to
resolve the issue.
66. I have been asked what I understood the cause(s) and consequences of
those issues to be. See paragraphs 62 and 63 above.
16 of 37
WITNO6060100
WITN06060100
67.1 have been asked what action was taken by POCL and ICL Pathway to
address those issues. Once the reality of the situation was understood by
both parties I believe there was a really strong effort on both sides to improve
the situation and work more closely to ensure that the data and the receiving
processes were more aligned and robust to issues arising. However, from
memory it felt that contractual factors meant that the joint endeavor on
reference data management was continually under pressure from a date
driven agenda by which I mean that the contractual acceptance process,
which was linked to payments to Pathway, was driven very aggressively by
Pathway who were seeking formal acceptance as early as possible.
68.1 have been asked if I consider undue pressure was placed on the Horizon
programme to commence and progress roll out before the system was shown
to be fit for purpose. Yes, I did feel that undue pressure was placed on POCL
to provide contractual acceptance and begin roll out.
69.As I recall, most POCL colleagues that were involved in the assurance and
acceptance process felt that undue pressure was placed on POCL to provide
contractual acceptance and begin roll out, both of which were linked to
payments to Pathway (I think).
70. I have been asked to describe the terms on which he Post Office proposed to
proceed with roll out of the Horizon System in January 2000. I don’t remember
the detail of this and from the documents relating to this section of your
questions I can find little to jog my memory, but from memory I believe that
there were several “patching” processes involved, particularly around the
17 of 37
WITNO6060100
WITN06060100
provision and processing of business data into the ICL technical domain and
the output of accounting data from ICL into the POL central accounts (TIP).
Additionally, I believe that there was to be a very high level of scrutiny around
errors arising from these two interfaces that should have been in place until
they were proven to be completely robust.
7
.I have been asked how ICL Pathway responded to those proposed terms. I
don’t remember the detail of this but my memory is that there was a marked
change in approach to working more closely at operational levels in order to
keep pace with the demands of the contractually driven schedule for roll out.
72.1 have been asked about my understanding of a number of issues on or
around 24'" November 1999 at the time of roll-out decision. I do not clearly
remember the detail of what my understanding was relating to the points
raised on or around 4" November 1999, however I believe that generally my
feelings were:
a. on the ability of the integrity control to detect all data errors - there were
doubts about this, hence the need for stress testing and temporary dual
checking at TIP;
b. on the level of disruption and cause of new incidents - I believe that this
was felt to be manageable within the constraints of the initial rollout
plan, assuming that the issues would eventually be resolved;
c. on POCL’s understanding of the design and operation of the emerging
reconciliation controls. See a and b above, I believe that all POCL
experts at the time believed that the controls were sufficient for an
initial step into rollout, assuming the future issues emerging would be
18 of 37
WITNO6060100
WITN06060100
resolved, and within the constraints of the contractual need to provide
agreement;
d. on the constraints on the level of testing of the emerging reconciliation
controls. I think at that tine POCL representatives believed that if the
situation worsened there would be further joint working with Pathway to
collect more data and resolve any issues arising.
73.1 have been asked what input I had in relation to the “Demand Position Paper”
produced by POCL in preparation for the checkpoint meeting in November
1999. I don’t remember this but probably very little, it would probably have
been authored by Ruth Holleran plus a number of business experts such as
Martin Box for the section on AI376.
74.} have been asked to describe the circumstances and conditions on which ICL
Pathway agreed to grant me access to the EPOSS End of Day High Level
Design. I don’t remember and can find no helpful reminders in the
documentation provided for this section of your questions but I imagine it
would have been a document review and walkthrough comprising a group of
ICL experts and POL experts. This would not have been my area of expertise
so I would have been very much led by TIP personnel and any technical
people present. The document would probably have been covered by an
NDA.
75.1 have subsequently been asked to consider the reference at page 3 of
POL00028541 to the grant of access to EPOSS end of day high level design.
19 of 37
WITNO6060100
WITN06060100
Having read the reference at page 3 of POL28541 I confirm the above
statement that I would have been at document review of the EPOSS end of
day high level design and that there would have been better informed
business experts at the subsequent workshop, which I may or may not have
attended.
76.1 have been asked about the outcome of the workshop to which reference is
made in POL00028541 at Item No.5. I don’t remember the workshop but see
paragraphs 74 and 75 of this statement.
77.1 have been asked to what extent ICL Pathway met the conditions for
proceeding with rol! out of the Horizon System in January 2000. I believe that
at the time it was felt that the contractual conditions had been met and that
was supported by legal advice at the time.
78.1 have been asked to consider POLO0090590 and FUJ00118186 which
records that “Both POCL and the Contractor acknowledge that at least one of
[the criteria in Parts A to C of Schedule 4 to the Second Supplemental
Agreement] was not met and accordingly that [ POCL’s right to postpone to
the resumption of roll-out from January, 2000] became exercisable” .I have
been asked in which respect (s) did ICL Pathway fail to meet Parts A to C of
Schedule 4 of the Second Supplemental Agreement. I assume that the failure
you refer to is the one highlighted by the email from Martin Box in the
documents you have provided.
20 of 37
WITNO6060100
WITN06060100
79.1 have been asked to clarify the nature of legal advice which I received
concerning ICL Pathway’s ability to meet the contractual conditions for
proceeding with roll out. I have no detailed memory of this. Our main business
expert for legal issues relating to the programme at that time was Keith
Baines who was supported by our external legal partners of the time,
Slaughter & May.
80.1 have been asked what limitations were identified in the reconciliation
controls which has been adopted to address Al 376. I don’t remember the
details but I believe that there was general unease within POL that although
specific issues related to incidents so far seen had remedies in place, further
issues due to flaws in the underlying code, or the interaction of business data
with automated processes, may lead to more issues arising and that therefore
additional monitoring and “work around” practices may be required, possibly
for some time.
8
.I have subsequently been asked by whom was this “general unease” about
cash account imbalances expressed. As stated elsewhere, I think most
colleagues involved in the issues relating to Al376 were uneasy that it had
been shown that new errors could occur for new, unexpected reasons. For
example I think Ruth Holleran and Martin Box shared this view.
21 of 37
WITNO6060100
WITN06060100
82.1 have also been subsequently asked what I understand the “work around”
practices to entail. These were, for example, additional checking at TIP which
was not a standard business process.
83.1 have been asked how these limitations were addressed. As above,
additional monitoring and procedures until confidence in the end-to-end
processes were achieved.
84.1 have been asked to describe the nature of my involvement in agreeing the
terms of the Third Supplemental Agreement. I don't remember but I imagine it
would be my usual role of reviewing documents and sharing opinions with
peers and superiors.
85.1 have been asked about the nature of my working relationship with Keith
Baines and Martin Box in January 2000. I don’t understand the question. Keith
Baines was the POCL lead person on the Pathway contract but had a good
working knowledge of the issues and often represented POCL on them. Keith
was senior to me. Martin Box was part of a core POCL business team
representing the TIP area and was a colleague of roughly the same grade as
me.
86.1 have been asked why my input was sought at this stage concerning the
terms on which POCL would be prepared to proceed with roll out of the
Horizon System in January 2000. I don’t know the answer to this and I don't
remember that being the case. As I have stated elsewhere, this was not my
22 of 37
WITNO6060100
WITN06060100
area of expertise. I would not have been able to understand or interpret the
contents of the error matrix included in POL00083897. I can only assume that
my name appears so predominantly because of my previous close working
with Pathway and because of my role in supporting Ruth Holleran.
87.1 have been asked what view I took of the terms on which POCL agreed to
proceed with roll out of the Horizon system. As stated elsewhere, I and close
colleagues such as Martin Box and Ruth Holleran were concerned that the
drive for contractual acceptance and fast roll out was too strong for the
seemingly unstable state of the system in some areas at that time.
88. In relation to FU00J118186 ...
a. I have been asked why certain types of cash account discrepancy were
excluded in January 2000 from the agreed error rate criterion of 0.6%. I
don’t know the answer to this, Martin Box would have most knowledge on
this and may remember the reason;
b. I have been asked to describe and explain the Error Matrix included at
Schedule 4 to the Third Supplemental Agreement. As stated elsewhere,
this is not my area of expertise in terms of the effect on TIP and I do not
know what effect these conditions may have had within the office, but I can
tell from the first page that these are EPOSS conditions, such as
successive rollover of a stock unit, which have presumably caused an
incident at the office or at TIP;
c. I have been asked about my understanding in January 2000 of the
“defences” which had been implemented by ICL Pathway in order to
23 of 37
WITNO6060100
WITN06060100
address ongoing cash account discrepancies. I don’t remember what my
understanding was at that time, and many of these error conditions would
have been unknown to me, but I can see from the first page of the
document that EPOSS conditions such as the one described above have
been addresses by “defences”, which in this case would presumably have
been an error message to the user and possibly prevention of cash
account rollover in the case of such error conditions being present;
. I have been asked what I understood the purpose of the “attribute checker”
to be. From the document I can tell that it is some code which predicts the
effect of reference data on the core system code and prevents the
reference data being accepted into the live domain to protect against
predicted (defined) error conditions. Presumably the term “attribute
checker” refers initially to the polarity (positive or negative) of the product
line;
. I have been asked about my understanding in January 2000 as to the
ability of the integrity control to detect all relevant data errors in Horizon.
As indicated above and stated elsewhere, it could only detect errors which
were known at that time and therefore posed a risk that other, as yet
undetected, error states may occur in future;
I have been asked what I understood ICL Pathway’s obligation(s) to be in
the event that it could not establish the root cause of a cash account
imbalance and/ or the appropriate corrective action to take to rectify the
imbalance. I don’t remember the detail of this. They may have accepted
that they must establish the root cause for the purposes of the contract but
in practice, any future such error conditions that occurred and for which
24 of 37
WITNO6060100
WITN06060100
the root cause could not be found, would surely have resulted eventually in
a declaration of “no fault found”.
89.I have been asked what I understood to be the effect of the amendments
made in Schedule 5 of the Third Supplemental Agreement to paragraphs
3.6.1.1 and 3.6.12 of Schedule G01 of the Codified Agreement. I don't
remember what my understanding was at the time and it’s difficult to
understand the effect now, reading the documents, without the contextual
understanding I had at that time. But I imagine that all POL reviewers
believed that the remedies proposed resolved the outstanding issues
sufficiently for POL to be contractually obliged to continue rollout so that ICL
Pathway could claim their revenues.
90. I have subsequently been asked to explain the basis of my statement in
paragraph 89. This is difficult to explain. The people at the sharp end of
assurance such as myself and Martin Box and many others were genuinely
doing their best and working very hard to ensure that the system was fit for
purpose. But there were many senior managers and legal people involved in
the higher decision making, and there was tremendous pressure from
Pathway and their legal team to agree to key milestones such as contractual
acceptance and rollout to achieve certain numbers that triggered financial
gains. It can be seen from the limitations in the additional contract schedules
that Pathway were (understandably) very persistent in closing down POCL
arguments for further assurance that would delay these milestones.
25 of 37
WITNO6060100
WITN06060100
91.1 have been asked if I consider that POCL was forced by the terms of the
contract to continue with roll out notwithstanding the general unease
about cash account imbalances to which t refer. Yes I do, see paragraph 90.
92.1 have been asked what consideration I gave at the time to the effect which
ongoing cash account imbalances would have upon SPMs, branch managers
and assistants who were obliged to implement and use the new Horizon
system. As I said above, the decisions that were made at this time were
exhaustively debated in POCL at the very highest level. As I recall, at that
time all departments were determined to put as many safeguards into
operational practices to protect the full POCL estate, including
subpostmasters, from a contracted out system and service management
regime that had some clearly recognizable issues.
93.1 have been asked to describe the conditions on which Al 376 was closed. As
above, the proposed remedies, additional monitoring and checking
procedures were deemed by POL representatives to be acceptable, at least
in terms of the contractual position at the time.
94. In relation to FUJ001 18186 and POL00028507 and paragraph 92 above ...
a. I have been asked to clarify what I understood the “proposed remedies,
additional monitoring and checking procedures" to entail. We have already discussed
the TIP integrity checks, the joint working on reference data, the attribute checker and
the contractual obligations accepted by Pathway to investigate new root causal issues
arising, at least at initial roll out stages. Mark Burley's email is a good example of others
that I may also have forgotten about: I dont remember what the “Miman‘ tool was, but it
would seem to be another defensive measure to issues arising, which were having
26 of 37
WITNO6060100
WITN06060100
to be developed reactively and against a commercial background of
Pathway insisting that changes should be at additional POCL expense;
b. I have been asked to clarify to whom “POL representatives” is referring.
Again, the email from Mark Burley is a good example: This refers to a
detailed assurance discussion that I was not party to between business
representatives (Dave Pye and Mark Burley) and Pathway representatives
(John Pope Pathway contractual, Steve Warwick Senior developer and
Phil Hemingway- possibly a reference data expert). From the outcome,
Mark is recommending closure of the Al (presumably 376) but he is
dealing with only a residual part of it and calling on other experts to
express an opinion. The eventual decision, whenever it came, would have
been a senior level one, relying on opinions from many players and
nuanced by the contractual position and a firm stance of closure from
Pathway;
c. I have been asked to clarify the nature of my role in the closure of Al 376.
As described above, mine would have been one of many opinions for
senior management to draw upon;
d. I have been asked if I consider that sufficient mitigation was in place to
I justify closure of Al 376. As described above and elsewhere, this was not
I a black and white decision and I would not have had a full picture.
27 of 37
WITNO6060100
WITN06060100
95.1 have been asked what concerns I had about the accounting integrity of
Horizon when roll out re-commenced in January 2000. See paragraph 80: I
shared the view that the system was not inherently robust and concerned that
the workarounds were required and may need additional strengthening as
more information about its performance became available through roll out.
96.1 have been asked if I expressed my concerns to others within POCL about
the robustness of the Horizon system in January 2000 and to whom. As
explained elsewhere, many of us in POCL shared this view, even through
and beyond the eventual closure of the original acceptance incidents and into
roll out.
97.1 have been asked who was responsible for monitoring the performance of
the Horizon system through its roll out. I believe that this responsibility
migrated to the service management team.
98.1 have been asked what role I played in monitoring the need for additional
strengthening of workarounds after roll out of Horizon recommenced in
January 2000. None that I remember.
99. I have been asked what action I took to report such concerns to others within
or outside POCL. The issue was generally discussed within POCL. I do not
I know whether the concerns were shared by others outside of POCL. I did not
raise them with any organisations outside of POCL.
28 of 37
WITNO6060100
WITN06060100
100.1 don't know if it is definitely t he case but I believe that concerns about the
robustness of the system, and the support service, were shared via senior
stakeholders and possibly through user groups.
101. I have been asked a number of questions about the processes and procedures
which were put in place to record and monitor the occurrence and cause(s) of cash
accounting data errors in Horizon. I was not involved with any of the implementation
of this. Although my name appears in the distribution list for document POL00043734
I was not involved with any detail and note that the same document distribution list
does not include me when the drafting was completed (POL00043741).
102. I have been asked what involvement I had in these activities. None, I was moved
to a very different role around this time and that may already have happened.
103.I have been asked when I ceased to be involved in matters concerning Al
376 and the cash account imbalances generated by the Horizon system. I
cannot remember the exact date. As I indicated earlier, my name appears on
draft 0.2 of the Pathway TPS reconciliation and incident management
document but I do not remember being invoived in this at all during live
running. Even at this early stage the author of the document is the Pathway
Service Manager. The POCL reviewers would almost certainly all have been
in the POCL service management domain or core business teams by the time
the document was approved.
29 of 37
104.I have been asked to which role within POCL I was transferred at this time. It
may have been the role I mentioned in paragraph 60. I don’t really remember
any of this clearly. The only clarity of recall in details such as that that I have
from this very distantly past is triggered by the documents you have
presented for my review.
105.Once the contractual acceptance process was completed the POCL service
management domain took over and would have been responsible, along with
Pathway service management, for ongoing management of known issues
and new issues arising.
IMPACT Programme
106.1 have been asked to describe the nature of my involvement in the IMPACT
programme as well as a number of questions about the programme more
generally. As far as I recall, none. The documents referenced here are early
versions. At that time I was a release manager and as such I am listed as the
assumed delivery manager for the S80 release. I think that the release was
eventually managed by Sue Harding as the vast majority of the content was
presumably Impact related. I did have involvement in the prompts and the
smart post projects where my role was akin to a business analyst. I can recall
no involvement in any of the detail relating to Impact.
General
30 of 37
WITNO6060100
WITN06060100
107.1 have been asked if, looking back, I consider that POCL effectively
scrutinised the trial and acceptance of the Horizon IT system. I think the
scrutiny was effective but contractual acceptance was very poor. POL were
very late in thinking through how this would happen. Until I moved back into
the core business I was in the joint working team and there was a very firm
screen between these two organisations resulting in core POL teams
effectively closing their eyes to the major event that was rapidly heading
towards them. Additionally, I believe the POL approach to contractual
acceptance was to pass it over to the joint team and rely on requirements
stipulating that Pathway were responsible for proving that the system was
acceptable. A lot of late effort went into preparing the receiving organization
for implementation but it was rather reactionary and events were being driven
principally by contractual deadlines which had very high commercial
consequences.
108.1 have subsequently been asked why I consider there as a “very firm screen”
between the joint working team and POCL’s core business in relation to the
issue of acceptance. I’m not sure about this. I think there may have been
some kind of contractual boundary. But I think the most likely problem in
practice would be that personnel in POCL core business areas were busy
with their day jobs and did not have spare capacity to consider the huge
change the programme would bring.
109.1 have been asked when I transferred from the joint working team back into
POCL's core business. I applied for a job, which I think would have been late
1998, possibly early 1999?
31 of 37
WITNO6060100
WITN06060100
WITNO6060100
WITN06060100
110.1 have been asked who was responsible for preparing the POCL core
business for acceptance of the Horizon system. I don’t know the answer to this.
111. [have been asked if! was concerned at the time that POCL had failed to
take adequate steps itself to assure the quality of the Horizon system. At the
time that I moved I don’t think I was aware that such major issues would arise.
112.I have been asked if! consider that the contractual deadlines for acceptance
were inappropriate. This is a very difficult question: I think I understood that
the contractual deadlines were necessary but it felt like the contractual
framework did not allow for the partnership between the host business and
the contractor to deliver the best outcomes.
113.1 have subsequently been asked what I meant by “very high commercial
consequences” in paragraph 107. As described elsewhere, I believe there
were large payments linked to the contractual acceptance milestone and the
roll out to a set number of offices.
114.have been asked what were the consequences applicable to ICL Pathway,
POCL or both. See above, I believe they affected both organisations.
115.I have been asked to explain what I consider should have been done to
make contractual acceptance more effective. I don’t feel I have enough
knowledge of the field to make a comment.
32 of 37
WITNO6060100
WITN06060100
116.1 have been asked if I consider the Horizon system was fit for purpose at the
point at which it was rolled out. I think it was as good as it could be given the
circumstances and the factors I have previously mentioned. Although the
system was widely considered to be suspect and had significant known
issues with accounting data, these had been thoroughly investigated by the
time mass roll out commenced and workarounds were in place. There was
also a lot of very close scrutiny of the performance of the system with regard
to usability issues and errors arising, with break points built in if required, as
far as I recall. What may not have been fit for purpose is the internal business
processes required to handle such a huge business transformation: I still
cannot understand how the Post Office management teams responsible
came to be so blind to the possibility that the system and the attendant work-
around procedures in place at the time could be responsible for accounting
errors, and failed so evidently in their duty to investigate them properly,
leading to such huge miscarriages of justice and such terrible human misery.
117.1 feel that something must have failed in the service management domain.
At hand over into live running the issues, including Al 376, were well known
and had been endlessly debated, including as far as I know, the risk
associated with going forward into roll out and the need for vigilance and
support. I do not understand how that position got so lost so quickly and,
moreover how the rest of the business were not aware that so many live
situations of misbalance were occurring.
33 of 37
WITNO6060100
WITN06060100
118.1 have been asked what steps were taken to notify these teams of internal
concerns about the robustness of the system and the reliability of the work-
around procedures which had been implemented by ICL Pathway and POCL.
See paragraph 116 above, as far as I know both POCL and Pathway service
management teams were well aware of the issues and the concems relating
to the robustness of the system.
119.l have been asked if I consider adequate feedback was obtained from end
users in relation to the reliability of the Horizon system prior to its acceptance
and roll out. I have no factual information for this question but I don't
remember seeing any strong evidence of it.
120.I have been asked if there are any other matters I consider will assist the
Chair. There are no further matters that I wish to bring to the attention of the
Chair at this time.
34 of 37
WITNO6060100
WITN06060100
Statement of Truth
I believe the content of this statement to be true.
Dated: 07-03. 222.3
35 of 37
Index to First Witness Statement of Graeme Seedall
WITNO6060100
WITN06060100
No. I URN
Document Description
Control Number
POL00030393
Electronic memorandum by the Post Office on
acceptance and business impact.
POL-0026875
In
POL00028342
Memo from Post Office re Acceptance
Workshop Roles
POL-0024824
lo
FUJ00118194
ICL Pathway Logical Design for
EPOSS/TIP Reconciliation Controls v0.8
22/12/99
POINQ0124358F
rs
FUJ00079178
Acceptance Proposal for Acceptance
Incident 376 (transmission of records to
TIP), 23 Sept 1999, version 0.9
POINQ0068766F
In
POL00028541
Letter from Tony Oppenheim, ICL
Pathway, to Keith Baines, POCL, re ICL
Pathway Response to POCL
Requirements for Rollout Decision
POL-0025023
Im
POL00090590
Email from Keith Baines to Jeff Triggs re
the latest stats for Al 376
POL-0090111
36 of 37
WITNO6060100
WITN06060100
FUJ00118186
POCL and ICL Pathway ‘Third
Supplemental Agreement
POINQ0124350F
loo
POL00083897
Fax from Jeff Triggs to Martin Box, POCL,
Chesterfield re: Schedules 4 and 5 of
Draft Suspension of Rollout Agreement
POL-0080881
Ko
POL00028507
Emails between Mark Burley, David
Stevenson, Min Burdett and Keith Baines
{all POCL) re Receipts Not Equal to
Payments
POL-0024989
POL00043734
ICL Pathway TPS Reconciliation &
Incident Management v0.2 3 Dec 99
POL-0040237
POL00043741
Draft ICL Pathway TPS Reconciliation &
Incident management Procedur Document
vi
POL-0040244
37 of 37