# RETRO-073-v1 — complete fictional source packet

Original educational fixture for the OpenMax AI sprint retrospective analysis article. Revised 2026-09-04. All items, notes, decisions and times below are invented. No actual people, customer data, OpenMax execution, measured product gain, benchmark or independent validation is represented. Scenario dates are January 2026; article verification is September 4, 2026.

## Scope and definitions

S14 starts 2026-01-05T09:00Z and ends at cutoff 2026-01-16T17:00Z. S15 starts 2026-01-19T09:00Z and ends at cutoff 2026-01-30T17:00Z. UTC throughout. Durations use calendar hours, not business hours. The period contains no example request withdrawals, repeated request episodes, reentries, export omissions or missing timestamps other than explicitly unresolved first responses.

Goal G14: administrators can review redacted ticket attachments with a recorded approval trail. Fictional G14-CHECK at 2026-01-16T16:00Z records successful permission-banner, redaction-preview and approval-trail checks for W01–W03. Goal met in this scenario; not a percentage of all selected items.

Completion checklist excerpt D14: applicable human review, automated checks, accessibility checks and supporting documentation must be complete. This is a scenario excerpt, not a universal or comprehensive Definition of Done. A “Done” board label does not override a failed applicable check.

## S14 work ledger

All W01–W08 were initially selected at 2026-01-05T09:00Z. W09/W10 were not. No other changes occur.

| ID | Item | Scope event after start | Retained at cutoff | Board state | D14 result and evidence |
| --- | --- | --- | --- | --- | --- |
| W01 | Permission banner | None | Yes | Done | Pass; D14-W01: applicable review, automated, accessibility and documentation checks recorded complete by cutoff |
| W02 | Redaction preview | None | Yes | Done | Pass; D14-W02: all applicable checks recorded complete by cutoff |
| W03 | Approval trail | None | Yes | Done | Pass; D14-W03: all applicable checks recorded complete by cutoff |
| W04 | Retry-state interface | None | Yes | Done | Pass; D14-W04: all applicable checks recorded complete by cutoff |
| W05 | Empty-state interface | None | Yes | Done | Pass; D14-W05: all applicable checks recorded complete by cutoff |
| W06 | Audit-search interface | None | Yes | In review | Not Done; B06 has no first review response at cutoff |
| W07 | Optional multilingual copy | Removed 2026-01-09T12:00Z | No | Backlog | Outside ending scope; not completed during S14 |
| W08 | Optional bulk-export keyboard support | None | Yes | Done | Not Done; D14-W08 applicable keyboard check fails; board status needs correction. No eligible first review request in this measurement cohort |
| W09 | Urgent retention fix | Added 2026-01-08T10:00Z | Yes | Done | Pass; D14-W09 all applicable checks complete; urgent maintenance excluded from normal-priority interface review cohort |
| W10 | Advanced filtering | Added 2026-01-12T10:00Z | Yes | In development | Not Done; no first review request by cutoff |

The status statements are invented source assertions for an arithmetic exercise, not real build logs or external quality verification. The response event is not itself evidence of final approval; D14 results are separate fictional assertions.

Reconciliation: initial8 − removed1 + added2 = ending9. Original-forecast Done W01–W05 =5/8=62.5%. Ending-scope Done W01–W05,W09 =6/9=66.666…%, displayed66.7%. Retained not Done W06,W08,W10=3. W07 remains in the original forecast denominator, but not the ending denominator. Do not label6/8=75% as original-forecast completion.

## Voluntary notes and export handling

Seven exported rows: N01-r1, N02-r1, N02-r1, N03-r1, N04-r1, N05-r1, N06-r1. The repeated N02-r1 is the same export object, not a second contributor. N05 is withdrawn before drafting, so no content is retained. There are6 stable note IDs and5 analyzable unique notes. No participant IDs are provided; distinct participant count cannot be computed.

| Note ID / revision | Approved fictional content or handling | Analysis status |
| --- | --- | --- |
| N01-r1 | W06 was still waiting for its first review at the cutoff | Perspective linked to B06; record supports the wait |
| N02-r1 | Try a reviewer rota so requests have an obvious destination | Proposal, not causal evidence; identical export duplicate removed |
| N03-r1 | W04 got a review response in one afternoon | Counterexample to “all reviews are slow”; supported by B04 |
| N04-r1 | W03 needed clarification of a test condition; a faster acknowledgment alone would not have completed its review | Participant perspective; B03 supports elapsed time, not this causal explanation |
| N05-r1 | Withdrawn; content not retained | Excluded from themes and approved text |
| N06-r1 | Same as last time | Insufficient specificity; optional clarification, not invented historical evidence |

Theme T01: some first review responses were delayed. Support B03/B06; counterexample B04; competing explanation N04. Routing/capacity and handoff clarity remain hypotheses, not proven causes. N02's proposal does not establish that the rota would solve them. No sentiment, blame, respondent count or consensus score is inferred.

## Review-response measurement contract

One repository, normal-priority interface parent work items. Eligible: first review request made within the Sprint and at least24 calendar hours before that Sprint's cutoff. Exclude urgent maintenance, removed items and items without a request. A substantive response is a human comment addressing the submitted change or asking a relevant clarification; acknowledgments and automated receipts do not qualify. A response is not necessarily approval or completed review.

For each eligible mature request, success means response_timestamp − request_timestamp ≤24h. An unresolved mature request stays in this denominator and is not a success. Open age is cutoff − request time, not eventual response duration. Any late request with less than24h follow-up must be reported as immature and revisited, not silently counted as failure or ignored. This example has no immature requests.

### Baseline S14 — six eligible requests

| Event ID | Item | Request UTC | First substantive human response UTC | Completed response hours | Open age at cutoff | Response within24h |
| --- | --- | --- | --- | --- | --- | --- |
| B01 | W01 | 2026-01-06T09:00Z | 2026-01-06T17:00Z | 8 | Not open | Yes |
| B02 | W02 | 2026-01-07T09:00Z | 2026-01-08T09:00Z | 24 | Not open | Yes |
| B03 | W03 | 2026-01-08T09:00Z | 2026-01-09T17:00Z | 32 | Not open | No |
| B04 | W04 | 2026-01-12T09:00Z | 2026-01-12T13:00Z | 4 | Not open | Yes |
| B05 | W05 | 2026-01-13T09:00Z | 2026-01-14T09:00Z | 24 | Not open | Yes |
| B06 | W06 | 2026-01-14T09:00Z | Unresolved | Not a completed duration | 56h | No |

Sorted completed-response durations:4,8,24,24,32. Median24h among5 responses. Timely numerator4, denominator6:66.666…%, displayed66.7%. Open requests1/6; age56h. All6 have a mature24h window.

### Follow-up S15 — six separate comparable items

F01–F06 are distinct normal-priority interface parent items in the same repository, not the same W01–W06 work repeated. Their scope is only the review-response cohort; this packet does not claim a complete S15 backlog or Sprint Goal assessment. Each has one eligible request. Item complexity has not been experimentally matched, so causal claims remain inappropriate.

| Event/item ID | Request UTC | First substantive human response UTC | Completed response hours | Open age at cutoff | Response within24h |
| --- | --- | --- | --- | --- | --- |
| F01 | 2026-01-20T09:00Z | 2026-01-20T13:00Z | 4 | Not open | Yes |
| F02 | 2026-01-21T09:00Z | 2026-01-21T13:00Z | 4 | Not open | Yes |
| F03 | 2026-01-22T09:00Z | 2026-01-22T17:00Z | 8 | Not open | Yes |
| F04 | 2026-01-23T09:00Z | 2026-01-23T17:00Z | 8 | Not open | Yes |
| F05 | 2026-01-27T09:00Z | 2026-01-28T09:00Z | 24 | Not open | Yes |
| F06 | 2026-01-28T09:00Z | Unresolved | Not a completed duration | 56h | No |

Sorted completed-response durations:4,4,8,8,24. Median8h among5 responses. Timely numerator5, denominator6:83.333…%, displayed83.3%. Open requests1/6; age56h. All6 have a mature24h window. Timely response change is one more request out of six,16.666…percentage points, displayed16.7points; not16.7% relative improvement. No statistical significance, causal impact or population estimate is claimed.

## X01 — trial and decision record

- Decision date: 2026-01-16T17:00Z.
- Hypothesis: explicit routing with staffed review coverage may improve first substantive response without transferring the burden outside agreed hours. Capacity and handoff quality are competing explanations to investigate.
- Owner in this invented case: rotating review coordinator. A real implementation requires a named person who accepts ownership, not only a role label.
- Intervention: morning triage and explicit reviewer assignment, 2026-01-19 through2026-01-30, within existing staffed hours. Do not bypass checks or pressure out-of-hours responses.
- Target: at least5 timely responses among6 eligible mature requests in this toy follow-up cohort. This is not a universal service objective; real targets need a population and operating context.
- Workload guardrail: zero additional review minutes outside agreed hours, assessed with aggregate workload log. No individual productivity ranking.
- Quality guardrail: zero newly identified critical defects attributable to the reviewed change within seven days after each release, using a severity definition agreed before the trial. The packet supplies no mature release-and-defect dataset and some required follow-up is not yet mature at Jan30; the gate is unknown and no defect-rate calculation is supported.
- Review cutoff: 2026-01-30T17:00Z.
- L14 baseline workload assertion:0 additional out-of-hours review minutes for this review process.
- L15 follow-up workload assertion:90 additional out-of-hours review minutes for this review process. These are invented aggregate log entries; no personal time records are provided.
- Response finding:5/6 timely meets toy target; one request remains open56h.
- Guardrail finding: workload fails; quality not yet established.
- Decision: ADAPT, not adopt. Stop out-of-hours escalation, arrange staffed reviewer capacity, then repeat the same measurement and complete the required quality follow-up. If capacity cannot be provided, stop the trial.
- Causal interpretation: unsupported. Work mix, staffing and other conditions were not controlled. This is a worked decision exercise, not evidence that the intervention or AI caused improvement.

## Replay checklist

1. Reconcile eight initial IDs, one removal and two additions to nine ending IDs.
2. Check5/8 and6/9 against their distinct populations; retain W08's failed check.
3. Remove only the repeated N02-r1 export; exclude withdrawn N05; retain N04 and the ambiguity of N06.
4. Subtract each timestamp pair, keeping unresolved ages separate from response durations.
5. Confirm the completed-only medians24h and8h with n=5 each, plus4/6 and5/6 timely responses and one open56h request each.
6. Inspect L15 and the immature quality window before making the decision. A draft that says “adopt” or claims an OpenMax performance gain has changed the meaning of the case.

This packet can support an offline arithmetic and draft-checking exercise. It cannot establish real facilitation quality, privacy compliance, tool reliability, customer outcomes or professional review.