WITNO05210200
WITN05210200
Witness Name: Ruth Reid
Statement No: WITN05210200
Exhibits: [ ]
Dated: 12 January 2023
POST OFFICE HORIZON IT INQUIRY
SECOND WITNESS STATEMENT OF RUTH REID
1. I, Ruth Reid, will say as follows:
2. In recognition of the importance of the Post Office Horizon IT Inquiry, I
continue to offer my full co-operation to the Inquiry.
Page I of 21
WITNO05210200
WITN05210200
3. 1am providing this statement following receipt of a second Rule 9 Request
dated 30 November 2022 from the Post Office Horizon IT Inquiry asking me to
provide a written statement regarding issues falling within Phase 2 and Phase
3 of the Inquiry. This Request contains 41 questions which the Inquiry have
asked me to address in my statement. For ease of reference, I have written
out each question with its corresponding question number.
PLEASE CONSIDER POL00083922 AND FUJ00000485.
Question 1: Please describe and explain how ICL Pathway proposed to rectify
Acceptance Incident 376 in early August 1999?
4. FUJ00000485 is a Change Control Note (“CCN”). I understand this to be a
contractual document between ICL Pathway Ltd (“Pathway”) and Post Office
Counters Ltd (“POCL”). I understand this CCN is the first of a number of
CCNs. I have not seen this document before. I was not involved in agreeing
contractual provisions. I believe this would have involved Dave Miller and
Keith Baines rather than myself.
5. POL00083922 contains a number of documents regarding a number of
acceptance incidents. As I have explained in my first statement, my work was
primarily focused on Acceptance Incident 376 (“Al 376”). This was a data
integrity issue between the electronic summary cash account stream and the
individual transactions which made up that cash account. Both streams of
data were produced by Pathway. The error was identified when comparing
these data streams with the information available to TIP (the Transaction
Information Processing project). This was designed as a temporary check to
ensure that TIP could receive a complete data set from Pathway.
Page 2 of 21
WITNO05210200
WITN05210200
6. I consider that it was worth highlighting that there continued to be a separate
supporting document stream produced by Transaction Process which fed the
business ledgers and the end to end error resolution process.
7. The document is produced by John Dicks of Pathway and confirms that the
incident was witnessed by Martin Box of POCL. The document confirms that
POCL consider the severity to be high as it “would affect POCL’s ability to
produce an accurate cash account”. The document goes on to confirm that
Pathway “accept the problem exists” but “would argue about the severity”. It
goes on to confirm that Steve Warwick (of Pathway) was “to provide
rectification of this issue. PWY understand the problem and are currently
working on the fix”.
8. As I do not have an IT and accounting background and was not on the End to
End Working Party focusing on this issue I cannot provide a technical answer
to the question. What I can say is that, to the best of my recollection,
Pathway proposed a change to the TIP interface to introduce the data integrity
control which was a control between the two data streams. To have that
introduced on their side would have required cross party resolution as part of
the business as usual end to end error resolution process, which was being
worked up by the End to End Working Party.
Question 2: Please explain why you considered that acceptance should be
declined in August 1999?
9. My recollection is that at this time the error rate was still 1.2%. Pathway
themselves were not content with this error rate at this time and therefore it
Page 3 of 21
WITNO05210200
WITN05210200
was agreed by everyone including myself that acceptance be declined in
August 1999.
Question 3: What (if any) concerns did you have at this stage about ICL
Pathway’s openness and willingness to share information and data concerning
the nature, severity and cause(s) of Acceptance Incident 376?
10. Due to the passage of time is it difficult to recall how I felt in August 1999 but
to the best of my recollection I do not recall having specific concerns at this
time. Both parties accepted that the error rate was too high and Pathway had
committed to continue work to reduce the error rate. Both parties committed
to a series of workshops to work on this and a number of other issues. I was
aware that there was a great deal of work to be done on both sides but there
was nothing to cause me to have concern regarding Pathway’s willingness
and openness to share information and data.
PLEASE CONSIDER POL00038332, POL00043711, POL00028466,
FUJ00118144, FUJ00118146 AND FUJ00118147
Question 4: What did you understand the root (cause)s of the accounting
discrepancies revealed by Acceptance Incident 376 to be?
11.Given that I do not have a technical / IT / accounting background I do not feel
able to provide an answer. However, I would say that Al 376 was considered
a data integrity issue rather than an accounting discrepancy.
Question 5: Did you consider that ICL Pathway had an adequate
understanding of the root cause(s) of the accounting discrepancies?
Page 4 of 21
WITNO05210200
WITN05210200
12.I consider Pathway adequately understood this data integrity issue as it
presented itself.
Question 6: How many outlets were affected by cash account discrepancies in
the summer of 1999?
13.1 don’t know.
Question 7: Did the control which you proposed be introduced by ICL Pathway
address the underlying root cause(s) of the accounting discrepancies?
14.My understanding is that the data integrity control was an enabler to identify
the reason for errors between the two Pathway produced data streams and
therefore would address the Al 376 data integrity issue.
15. Following the introduction of Horizon, there was still the need for error
resolution by the Transaction Processing Business Unit in Chesterfield and it
was accepted that we still needed a process to manage accounting
exceptions where transaction were not automated at the point of sale/data
capture after the developers handed over the Horizon system.
Question 8: If not, what action did you understand ICL Pathway to be taking to
address the root cause(s) of the accounting discrepancies?
16. Pathway were responsible for resolving the Al 376 data integrity issue as set
out above. Pathway were not responsible for all POCL accounting
discrepancies.
17.In terms of the action Pathway were taking to address the data integrity issue
at this time document POL00028332 sets out the work Pathway have done to
identify the error and contributory problems and fixes developed and
Page 5 of 21
WITNO05210200
WITN05210200
implemented as at 12" August 1999. Document POL00043711 contains the
minutes of subsequent meetings on 25" and 26" August 1999 which again
provide an account of the progress made and actions required. To the best of
my recollection these documents accurately set out the action Pathway were
taking.
Question 9: Did you consider that the action being taken by ICL Pathway to
investigate the underlying root cause(s) of the accounting discrepancies was
sufficient? Please explain the reason(s) for your answer.
18.Yes, for the reasons set out above and having regard to the actions Pathway
were taking as set out in the documents referred to.
PLEASE CONSIDER FUJ00079176 AND FUJ00118150
Question 10: Please describe and explain how the error rate of 0.6% came to
be proposed by you and Keith Baines as a condition for acceptance of the
Horizon system?
19. There was an End to End Working Party which came up with the 0.6% error
rate which came to be proposed by Keith Baines and I. This Working Party
included Mark Burley who worked for me and various members of Keith
Baines’ team. For the avoidance of doubt, this working party was not confined
to those with technical/IT backgrounds but also included Transaction
Processing, Finance, the Operational Retail Network Management and the
Investigation and Audit teams. I believe that the 0.6% error rate would have
had to have been agreed by the Head of Transaction Processing, which was
the relevant business unit. This position owned the operational business
process, which was the relevant school of management thought at that time.
Page 6 of 21
WITNO05210200
WITN05210200
20.1 was the representative of these group at the Acceptance Workshops and
therefore the conduit for feeding into Pathway, the findings of the End to End
Working Party.
Question 11: Why did you consider such an error rate to be acceptable?
21.1 believe it would have been the position that if an error rate of 0.6% was
acceptable to the Head of Transaction Processing Business Unit, the
technical representatives of the relevant working group and Keith Baines (as
POCL Commercial Lead), with the impact of accepting such an error rate
having been assessed and costed, I considered it acceptable.
Question 12: What did you understand the impact of such an error rate to be
on the network of post office branches once the Horizon system had been
rolled out nationwide?
22.As I mentioned above, agreeing an error rate of 0.6% would have been
subject to an impact assessment which would have considered the cost and
resource required to investigate and resolve the residual errors. There was a
Business as Usual process for dealing with errors, which existed prior to
Horizon because errors were a fact of life where transactions were not fully
automated and required manual input of information at the outlet. That
remained the position following the roll out of Horizon.
23. There was a large number of people in the Transaction Processing Business
Unit in Chesterfield whose role it was to investigate and resolve errors. As a
result, I considered that the network of Post Office branches would have been
supported by these business processes in managing any residual errors
resulting from Al 376 data integrity issues.
Page 7 of 21
WITNO05210200
WITN05210200
24.For the avoidance of doubt the End to End Working Party were responsible
for augmenting the end to end business resolution process as a result of
introducing both Horizon and the data integrity check.
Question 13: How did you envisage that POCL and ICL Pathway would manage
the resolution of errors at a rate of 0.6% nationwide?
25. Following roll out the resolution of errors would have been the role of the
Transaction Processing Unit in Chesterfield following the Business as Usual
process with input, as required, from the Retail Network, Finance, Internal and
External Audit and the Internal Investigation teams. At this time, there was a
team of around 700 people in Chesterfield within Transaction Processing
involved in this process.
Question 14: Please describe and explain the other conditions which you
proposed as part of the resolution of Al 376 prior to acceptance?
26.1 believe the question refers to the 6 conditions which are referred to at page 9
of document FUJ00079176 as being proposed by Keith Baines and I. There
is also a reference to a document called “RH’s paper” on page 9 of that
document. I have not seen that paper and I cannot recall from memory what
the other conditions were. These conditions and the resultant papers, were
produced by the End to End Working Party.
PLEASE CONSIDER POL00038907.
Question 15: Who was responsible for determining whether or not acceptance
of the Horizon system should take place in September 1999?
Page 8 of 21
WITNO05210200
WITN05210200
27.To the best of my recollection this would have fallen to the Release
Acceptance Board. I do not know if there was one individual who had ultimate
responsibility. For the avoidance of doubt, I confirm that I was not a member
of the Release Acceptance Board and responsibility did not sit with me.
Question 16: At paragraph 92 of your statement dated 13 October 2022, you
State that Peter Copping, an Independent Expert, recommended “as part of the
contract arbitration process, on the basis that there was a rectification plan to
introduce the data capture release in December 1999 (this was a large piece of
work), that acceptance take place but on a conditional basis, rather than
acceptance continue to be declined.”
a. Please confirm how and when this recommendation was
communicated to you and/or to POCL.
28.1 do not specifically remember how it was officially communicated to POCL
but from my perspective, my own knowledge of Peter Copping’s
recommendations came from the discussions which took place at the
acceptance workshops at the Renaissance Hotel at Heathrow.
b. Please confirm what (if any) discussions you held with other
employees of POCL concerning this recommendation.
29.1 do not specifically remember but I would have discussed it with Dave Smith,
who was my boss. I would be talking to him daily or several times a day and
we would talk about everything all the time, before and after every meeting.
He was based in Gavrelle House in London. I know I also spoke with Mark
Burley and Peter Jones on a daily basis and it is very likely I would have
therefore discussed it with them. The same can be said for the Automation
Page 9 of 21
WITNO05210200
WITN05210200
Working Group which I was a member of - it is very likely that I would have
discussed it with my colleagues on the AWG, in particular Sue Harding of
Transaction Processing, Shaun Delaney of POL Finance and Carol Pryslak
and one other person (I do not remember their name), who represented the
Regions of the Operational Retail Network on the AWG. Dave Smith would
have represented this type of issue at the Automation Programme Group
level
c. Did you consider this to be a binding recommendation by the
independent expert? Please explain the reason(s) for your answer.
30.1 do not know the contractual or legal status of Peter Copping’s involvement
and therefore I do not know whether his recommendations were binding.
Question 17: Why did you prefer for the data integrity control to be
implemented prior to acceptance?
31.From a common sense viewpoint I thought it was better for Pathway to
implement the control prior to formal acceptance rather than to rely on POCL
through the TIP process and because the TIP checking process was not
designed to be a permanent process. I wanted to see it fixed and the target
end to end operational process in place before acceptance.
Question 18: What (if any) risks did you perceive there to be in granting
conditional acceptance to the system prior to implementation of the control?
32.To the best of my recollection, conditional acceptance wasn’t something I had
considered and I didn’t know it was possible. As far as I was aware, the
concept of conditional acceptance was introduced/raised by Peter Copping
and the respective commercial teams. Once I understood how conditional
Page 10 of 21
WITNO05210200
WITN05210200
acceptance could work, and what the conditions could/would be, I was
satisfied with the recommendation and did not perceive any particular risks on
the basis that Pathway would still be required to demonstrate an improvement
to the system to bring the error rate within acceptable limits in order for
acceptance to take place.
Question 19: To whom (if anyone) did you communicate your preference and /
or concerns about the risks of granting conditional acceptance to the system?
33.1 would have discussed it with my colleagues as outlined in my answer to
questions 16b. above.
Question 20: To your knowledge, who ultimately took the final decision in
respect of acceptance of the Horizon system?
34. Please see my response to question 15 above.
PLEASE CONSIDER FUJ00079316, FUJ00118143 AND FUJ00118194
Question 21: What (if any) involvement did you have in scrutinising the design
for the EPOSS / TIP Reconciliation Controls?
35. The technical work of scrutinising the design for the EPOSS / TIP
Reconciliation Controls would, to the best of my recollection, have been
carried out by the End to End Working Party. I don’t recall ever seeing any of
the Pathway design documents. I do not know whether such document would
have been shared with anyone on the POCL side involved in the technical
work.
Question 22: What did you understand about how these controls were
intended to function?
Page 11 of 21
WITNO05210200
WITN05210200
36.As I do not have a technical/IT/client accounting background this is not within
my knowledge.
Question 23: Were you satisfied that these controls were sufficiently rigorous
to address the accounting integrity issues identified y Al 376? Please explain
the reason(s) for you answer.
37. Again, I am unable to provide an answer from a technical/IT perspective. The
implementation of data integrity controls was focused on ensuring that the
error rate of the Al 376 data integrity issue was reduced to 0.6% or less,
otherwise Core System Release would not take place.
Question 24: Who within POCL was responsible for investigating and
resolving the manual error reports produced by ICL Pathway pursuant to these
controls?
38. The Transaction Processing Business Unit in the first instance. Depending on
the nature of the issue there may have been a need for involvement from the
Operational Retail Network Management, the Finance Executive, Pathway
and the Investigations Department.
Question 25: How were accounting discrepancies caused by POCL Reference
data intended to be resolved?
39.To the best of my recollection, this would have been the work of the
Reference Data Project Team (Ted Baldwin and Geoff Darby) in conjunction
with John Meagher of the Horizon Product Assurance Team. I was not part of
either of these teams.
PLEASE CONSIDER FUJ00118176 AND FUJ00118171
Page 12 of 21
WITNO05210200
WITN05210200
Question 26: Are these the types of daily and weekly reports to which you refer
in paragraph 112 of your witness statement dated 13 October 2022?
40.No. For an example of the reports I was referring to see page 2 of document
POL00028559.
PLEASE CONSIDER POL00043723.
Question 27: At the time of the roll-out decision on or around 24 November
1999, what was your understanding of the following:
a. The ability of the integrity control to detect all data errors;
41.1 can only rely on what is contained within document POL00043723 to assist
me in providing an answer to this question. POCL’s position was that the
Pathway integrity control was capable of detecting all relevant incidents. This
is not the same as having no errors but what was required, as the document
confirmed, was a robust process for error detection of the two data streams.
My recollection is that the error rate must have been reduced to 0.6% or
under which was what had been previously required.
b. The level of disruption and cause of new incidents;
42.1 am unable to recall and the documents I have reviewed do not assist me in
providing an answer.
c. POCL’s understanding of the design and operation of the emerging
reconciliation controls;
43.1 am unable to provide an answer regarding design as I do not have a
technical/IT background. I understand the purpose/operation was to enable
the data streams to be better compared and for errors to be identified and
Page 13 of 21
WITNO05210200
WITN05210200
reported so that they could then be investigated by the Transaction
Processing Business Unit.
d. The constraints on the level of testing of the emerging reconciliation
controls.
44.To the best of my recollection, whilst there remained a lot of work to do on
both sides, I believe that we were satisfied that there was sufficient time and
resource to carry out testing of the emerging reconciliation controls. This
factored in the operational impact of the Christmas period and an accounting
period end.
Question 28: What (if any) input did you have in relation to the “Demand
Position Paper” produced in preparation for the checkpoint meeting in
November 1999?
45.1 don’t recall having any input. The document is unhelpful as it does not
contain details of author(s) and distribution. It also appears that the document
is perhaps two versions of the same document or two separate documents,
with pages 1-4 largely repeated on pages 5-8. I don’t remember seeing it
before.
Question 29: What action did you understand POCL to be taking to address
the ongoing concerns about cash account discrepancies in November 1999?
46. The work of the End to End Working Party continued throughout this period.
Due to the passage of time I cannot remember any specific details.
Page 14 of 21
WITNO05210200
WITN05210200
PLEASE CONSIDER FUJ00118186
Question 30: Why were certain types of cash account discrepancy excluded
from the agreed error rate criterion of 0.6% (as originally defined in the Second
Supplemental Agreement) in January 2000?
47.1 do not feel that I am in a position to comment on the content of the Second
and Third Supplemental Agreement as these are contractual documents and I
was not involved in the drafting or negotiation of the contract between
Pathway and POCL. I don't recall ever seeing these documents.
48.FUJ00118186 is a document signed by Dave Miller in the presence of Keith
Baines and dated 19" January 2000. As I explained in my first statement, I
don’t recall being present in January 2000 and I often took an extended
holiday in January.
49.1 believe the contract would have reflected the fact that the agreed error rate
would only point to exceptions between the individual transactions stream and
the cash count transaction stream as compiled by the Horizon system. There
were many different causes of errors not all of which were attributable to
Pathway, for example those attributable to human error at the point of input of
information at an outlet. Many of our clients’ transactions were not
automated. Automated products were known within the business as ‘token
drive’ products. I believe that the contract would have recognised that
Pathway weren’t responsible for errors which derived from other legacy
processes/systems/products which hadn't been fully automated/were not
token driven and which required manual input.
Page 15 of 21
WITNO05210200
WITN05210200
Question 31: Please describe and explain the Error Matrix included at
Schedule 4 to the Third Supplemental Agreement.
50.As I wasn’t on the End to End Working Party, all that I feel able to comment
on in relation to Schedule 4 is that these are references to the type of errors
found at the TIP interface as examined by the Working Party.
Question 32: What was your understanding in January 2000 of the “defences”
which had been implemented by ICL Pathway in order to address ongoing
cash account discrepancies?
51.The term or word “defences” isn’t something I recall being used during my
work on these matters. The three defensive measures referred to at note no.
6 (page 24) of document FUJ00118186 relate to technical matters which I am
unable to comment on.
Question 33: What did you understand the purpose and function of the
“attribute checker” to be?
52. Again, this is not something that I recall being aware of at the time and the
document doesn’t help me to understand or recollect what the “attribute
checker” was.
Question 34: What was your understanding in January 2000 as to the ability of
the integrity control to detect all relevant data errors in Horizon?
53.As I have mentioned above and in my first statement, I do not remember
being involved in January 2000, but I can only assume that all parties were
satisfied that the data integrity control was capable by this stage of detecting
all relevant data errors in Horizon.
Page 16 of 21
WITNO05210200
WITN05210200
Question 35: What did you understand 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?
54. My understanding is that Pathway were responsible for issues/errors caused
by their system and would therefore need to provide a fix for any issues which
were attributable to them.
PLEASE CONSIDER FUJ00080690
Question 36: Were you shown a copy of this report at or around the time it was
first produced in September 1998?
55.1 do not recall seeing this document. The author names do not mean anything
tome. This was a Pathway only report as far as I can see.
Question 37: Were you aware, at the time, of the concerns expressed by ICL
Pathway developers within this report concerning the quality of the EPOSS
application, code and design?
56.No.
Question 38: What (if any) impact would knowledge of these concerns have
had upon your handling of Acceptance Incident 376 had you been aware of
them at the time?
57. This report appears to contain concerns attributable to the authors regarding
matters between August and September 1998. I feel that it is impossible for
me to say, had I know about this, whether this would have had any bearing on
the process of dealing with Acceptance Incident 376, which was of course
being worked on constantly from the Summer of 1999 onwards.
Page 17 of 21
WITNO05210200
WITN05210200
PLEASE CONSIDER FUJ00079782.
Question 39: Were you shown a copy of this audit report at or around the time
it was produced in October 1999?
58.No.
Question 40: Were you aware, at the time, of the concerns expressed by ICL
Pathway auditors within this report concerning the quality of the EPOSS
application, code and design?
59.No. I did not have any interaction with Pathway auditors.
Question 41: What (if any) impact would knowledge of these concerns have
had upon your handling of Acceptance Incident 376 had you been aware of
them at the time?
60.1 did not receive this document but note that it does highlight a number of
issues with the ICL Pathway system/code. As I do not have a technical/IT
background it is difficult for me to comment on whether these issues,
contained in a report which is a snapshot view from September 1999, are
significant in terms of the subsequent Core System Release which took plan
in January 2000. It is now 2023 and I feel that it is impossible for me to say,
had I know about this over 23 years ago, whether this would have had any
bearing on the process for resolving Al 376. It was Pathway and not POCL’s
responsibility to develop their system in accordance with the requirements of
private finance.
Page 18 of 21
WITNO05210200
WITN05210200
Statement of Truth
I believe the content of this statement to be true.
Signed:
Dated: 12-1-23
Page 19 of 21
Index to Second Witness Statement of Ruth Reid
WITNO05210200
WITN05210200
URN
Document Description
Control Number
POL00083922
Fax from Keith Baines to Jeff
Triggs re Incident Reports
(Horizon)
POL-0080906
In
FUJ00000485
Supplemental Agreement
(POCL & ICL Pathway),
Change Control Notice 550, 20
August 1999 re CSR
Operational Trial Review
Period
POINQO006656F
loo
POL00038332
PO: Additional End of Day
Handout, v2
POL-0035267
s
POL00043711
Email from Altea Walker to
Graeme Seedall and others, re
agenda for 2 Sep acceptance
workshop, with enclosures
POL-0040214
lo
POL00028466
Acceptance Proposal for
Acceptance Incident 376 - not
passing records to TIP due to
harvester exceptions caused
by missing functions in counter
code
POL-0024948
In
FUJ00118144
TIP Incident 376 Status Report
2/09/99
POINQ0124308F
IN
FUJ00118146
Al 376 - TIP Incident Status
Report 16/09/99
POINQ0124310F
loo
FUJ00118147
Al 376 - TIP Incident Status
Report 22/09/99
POINQ0124311F
1
FUJ00079176
Acceptance Workshop (7) -
Action Points, 17 Sept 1999
POINQ0068764F
FUJ00118150
ICL Pathway Acceptance
Resolution Plan Acceptance
Incident 376 v0.9 dated
23/9/99
POINQ0124314F
POL00038907
Impact Programme - POL FS
to SAPADS - Application
Interface Specification (Issue
No: 0.2)
POL-0035389
FUJ00079316
Agreement between POST
OFFICE COUNTERS LTD and
ICL PATHWAY LIMITED
(Second Supplemental
Agreement CCN)
POINQ0085487F
FUJ00118143
Emergency CCN 562 settling
out new paragraph 3.6 to be
POINQ0124307F
Page 20 of 21
WITNO05210200
WITN05210200
added to Schedule G1 to the
Codified Agreement
14 =I FUJ00118194 ICL Pathway Logical Design POINQ0124358F
for EPOSS/TIP Reconciliation
Controls v0.8 dated 22/12/99
15 I FUJ00118176 _I TIP Incident Table POINQ0124340F
416 I FUJ00118171 Post-Acceptance TIP Incident I POINQ0124335F
Status Report - 18/11/99
Az POL00043723 Roll-out Decision - Demand POL-0040226
Position Paper (undated)
18 I FUJ00118186 =I POCL and ICL Pathway ‘Third I POINQ0124350F
Supplemental Agreement’
419 I FUJO0080690 ‘I Report on the EPOSS PinICL I POINQ0086861F
Task Force
20 FUJ00079782 ICL Pathway CSR+ POINQ0085953F
Development Audit v1
Page 21 of 21