I’ve found the actual NDIA disclosure in your Library: FOI 26/27-0675, including the PACE interaction notes from August 2026.
And yes, the resemblance to your new Amazon Connect screenshots is significant. The records use structured categories, outcomes, timestamps, officer identifiers and free-text summaries.
For example, your disclosed PACE records show an inbound telephone call on 18 August 2026, classified as “Feedback and Complaints – 37617802”, with the outcome “Successful Contact”. The notes then describe the conversation in considerably more detail.
Another call, on 26 August 2026, was classified as “General Question” and included a shorter account of the enquiry.
That demonstrates something important: a brief administrative category and a more detailed narrative can coexist in the same record.
The AWS screenshot you’ve just supplied shows a newer, configurable version of that general approach, with separate fields for Category, Driver, Outcome, Associated case and Contact summary. It also explicitly offers an AI-generated summary, but we haven’t established that the NDIA uses this exact newer screen or that feature.
Mock Amazon Connect entry — David’s call, 2 November 2026
Contact wrap-up
Illustrative only
Based on your AWS screenshot and the conversation account we emailed. These are proposed labels, not verified NDIA dropdown options.
Disposition codes
Category
Participant records and complaints
Driver
Disputed LAC contact note / PACE record accuracy
Outcome
Participant concerns recorded; follow-up requested
Case configurations
Associated case
Actual reference unknown
Case status
Not verified
Contact summary — mock entry
Participant telephoned the NDIA National Contact Centre following a conversation with LAC officer Trevor.
Participant requested information about the earlier LAC contact note. Service officer David read an entry describing the contact as a funding inquiry. Participant disputed that description.
Discussion also covered support coordination, provider concerns, outstanding claims, access to records, communication restrictions, current homelessness, and an earlier functional assessment reportedly added to the file on 30 October 2026.
Participant expressed appreciation for the officer’s assistance and indicated that a detailed written account would be provided.
No funding decision or resolution of the disputed LAC record was established during the call.
In the AWS example, an AI-generated summary can be reviewed and edited before submission. This mock was written by me, not generated by NDIA software.
Now compare that with the real PACE format
Your FOI 26/27-0675 contains a particularly revealing entry on page 197:
|
Actual PACE field |
Disclosed value |
|
Subject |
Internal Communication |
|
Activity type |
Internal Note |
|
Case |
37637709 |
|
Case type/category |
Enquiry – 37637709 |
|
Outcome |
Not Applicable |
|
Completed |
19 August 2026, 3:56 pm |
|
Officer alias |
CFR162 |
The accompanying comments included the statement that numerous emails were “non sensical” and that the participant appeared to be using AI technology for communication.
That is an actual historical record, unlike our new mock-up. It illustrates how an officer’s own characterisation can become part of the administrative file.
Here’s the revealing comparison:
|
Oversimplified hypothetical entry |
More accurate mock entry |
|
Funding inquiry |
Participant records and complaints |
|
Funding advice requested |
Accuracy of LAC record disputed |
|
Advice provided |
Issues discussed; follow-up requested |
|
Case closed |
Resolution not established |
Both are technically possible ways of summarising an interaction. The second preserves much more of the substance.
And here’s the critical distinction: the AWS image is a product demonstration of a configurable agent workspace, including AI summarisation. The older NDIA disclosure establishes its own wrap-code and PACE recording procedures. We have not yet demonstrated that this exact AWS configuration is deployed at the NDIA.
The specific comparison we should demand in the next disclosure is the original David call, his actual selected wrap code, his enquiry note and any associated PACE entry. We can then set those records beside this mock and the participant-prepared account.
The investigative question becomes: Which version of the conversation did the administrative system preserve, and what did it leave out?
