# AtlasOps Cloud v4.2: Source-Linked Improvement Backlog Real request records were supplied, so the built-in demonstration was not used. ## 1. CONTEXT AND ADMISSION **Domain profile (attributed to the supplied fields)** | Item | Value | |---|---| | Sector | B2B enterprise workflow software (PRODUCT_SCOPE) | | Product/version/deployment | AtlasOps Cloud v4.2, EU production | | Jurisdiction/applicability | The "EU" deployment label is supplied. Applicability of any specific law is **unknown**; no legal conclusion is drawn. | | Direct users | Reporting users (shared cohort of 500) | | Indirect beneficiaries | Unknown | | Outcome | Reduce manual reporting work for eligible business users | | Observation window | January–September 2026. Cutoff date unknown. | | Future horizon | Calendar 2027 | | Operating calendar, actual integrations | Unknown | | Accountable owners | Product security lead is named as reviewer (EV-106). Accountable owners for each gate are not otherwise established. | **Shared scoring policy (POL-101)** - R is the selected affected-user forecast. - I is the supplied shared-scale judgment. - C is one supplied factor. - E is total cross-functional person-months. - G = R·I·C is a heuristic priority index, not money or measured benefit. - Impact multipliers are heuristics, not observed effects. **Admission** | Record | Status | Reason | Locator | |---|---|---|---| | REQ-101 | Admitted | Future improvement need (dashboard export) | REQ-101/description | | REQ-102 | Admitted | Future improvement need (saved filters) | REQ-102/description | | REQ-103 | Admitted as lineage only | Repeat of REQ-101 by the same originator for the same need. Not independent demand. | REQ-103/follow-up | | REQ-104 | Excluded | Outage/restoration content. No demand or RICE contribution. | REQ-104/description | | REQ-105 | Excluded | Routine provisioning. No separate improvement need is supported, so no hypothesis is ranked. | REQ-105/description | No parent records are present. **Duplicate group:** {REQ-101, REQ-103} forms one demand instance. **Corpus totals:** 5 raw records, 3 admitted (2 demand instances), 2 excluded. Supporting evidence (EV-101 to EV-106) is non-demand context. ## 2. THEMES AND REQUIREMENTS | Theme/candidate ID | Raw request IDs/count | Proposition IDs/count | Demand instances/membership | Source authors | Requesters/count | Originators/count/standing | Requested-for/count | Business units/count | Dates/original ID-field locators | Observation window | |---|---|---|---|---|---|---|---|---|---|---| | C1 Dashboard export | REQ-101, REQ-103 / 2 | P-101 / 1 | 1: {REQ-101, REQ-103} | Maria Chen (both) | Unknown / unknown | Maria Chen / count 1 / reported (repeat annotated as same originator; independence beyond that unknown) | Unknown / unknown | Finance Operations / 1 (stated beside the submitter) | 2026-06-12 (REQ-101/description); 2026-07-19 (REQ-103/follow-up) | Jan–Sep 2026 | | C2 Saved report filters | REQ-102 / 1 | P-102 / 1 | 1: {REQ-102} | David Miller | Unknown / unknown | David Miller / count 1 / reported | Unknown / unknown | Operations / 1 (stated beside the submitter) | 2026-07-03 (REQ-102/description) | Jan–Sep 2026 | **C1 Dashboard export** - **Source quotation:** "Allow authorized reporting users to export the fields visible in their dashboards. We currently copy values manually into spreadsheets." (REQ-101/description). REQ-103/follow-up repeats the need. - **Need:** Remove manual copying of dashboard values. - **Scoped gap/counterevidence:** EV-101 reports that standard reporting users cannot export while administrators can. EV-101 does not state version or deployment, so a match to v4.2/EU production is provisional. No intent or root cause is established. The administrator export is partial counterevidence for the underlying capability, but it applies to a different role. - **Candidate requirement (proposed):** Authorized reporting users can export the fields they can already view. - **Proposed acceptance criterion:** An authorized reporting user exports exactly the fields visible on their dashboard, and a user without authorization cannot. This is proposed and has not been tested. **C2 Saved report filters** - **Source quotation:** "Let users save report filters between sessions so they don't need to rebuild the same reports." (REQ-102/description) - **Need:** Avoid rebuilding filters each session. - **Scoped gap/counterevidence:** EV-102 is a supplied project test note, not independent verification, and it reports that filters are not persisted. It does not state version or deployment. No root cause is established. - **Candidate requirement (proposed):** A user's report filters persist across sessions. - **Proposed acceptance criterion:** After ending and restarting a session, the user's previously applied filters are present without re-entry. This is proposed and has not been tested. ## 3. CANONICAL RICE Both candidates share the outcome, the 500-user cohort, the CY2027 horizon, the Impact scale and the person-month effort unit, so they are one comparison. Overlap between the two reach sets is unknown, so no reaches are summed. | Rank | Candidate | R | I | C | E | G | RICE | Factor provenance | |---|---|---|---|---|---|---|---|---| | 1 | C2 Saved filters | 300 (P=500 eligible × A=0.60; cohort 500) | 1 | 0.8 | 4 | 240 | 60 | R: EV-103, forecast, approved, CY2027, date not supplied. I: EV-104, judgment, approved, date not supplied. C: EV-104, supplied single factor, judgment, no components. E: EV-104, approved estimate, 4 person-months (internal development, integration, testing, rollout), date not supplied. | | 2 | C1 Export | 150 (P=300 eligible × A=0.50; cohort 500) | 2 | 0.8 | 5 | 240 | 48 | R: EV-103, forecast, approved, CY2027, date not supplied. I: EV-104, judgment, approved, date not supplied. C: EV-104, supplied single factor, judgment, no components. E: EV-104, approved estimate, 5 person-months (same scope), date not supplied. | **Checks:** R = P×A was verified for both (300×0.5=150; 500×0.6=300). No adoption factor was applied twice. **Unranked evidence list:** None. Both candidates have a complete factor set. **Effort caveats:** EV-104's effort scope does not mention vendor labour or security/privacy assurance. Those components are unknown, not zero. **Leader comparison:** The two candidates have equal G, so the tie in G does not decide the order. C2 has the lower E (4 vs 5). Cross-product check: G_C2 × E_C1 = 240×5 = 1200, versus G_C1 × E_C2 = 240×4 = 960. 1200 > 960, so C2 leads. The lead comes entirely from lower effort, not from Reach or Impact. **Boundary.** Vary only C2's effort, with T as the symbol for the threshold: T = (R_C2·I_C2·C_C2) / ((R_C1·I_C1·C_C1)/E_C1) = (300×1×0.8) / ((150×2×0.8)/5) = 240 / 48 = **5** (exact, not rounded). Substitution check: at E_C2 = 5, C2 scores 240/5 = 48, which equals C1's 48. | C2 effort | Result | |---|---| | Below T | C2 leads | | Exactly T | Tie | | Strictly above T | C1 leads | This is sensitivity arithmetic, not an effort forecast. ## 4. GATES, DISPOSITIONS AND DECISION BRIEF | Item | C1 Export | C2 Saved filters | |---|---|---| | Safety gate | **UNKNOWN**: EV-106 says the scoped review is not completed. The date is not supplied. | **UNKNOWN**: same source. | | Compliance/privacy gate | **UNKNOWN**: same source. Applicability of any EU law is not supplied. | **UNKNOWN**: same source. | | Gate owner | Product security lead is named as reviewer. Accountable owner and intended-use statement are not supplied. | Same. | | Closure condition | A completed, scoped, dated review of the intended use and EU production scope, with a named accountable owner. | Same. | | Delivery eligibility | **HOLD** | **HOLD** | | Primary disposition | **Further discovery — proposed** | **Further discovery — proposed** | | Basis | EV-105 says the feature is technically possible but no implementation approach is validated (POL-104). | Same. | | Unconfirmed alternatives | Permission/role configuration change; native export feature; integration with an existing tool. None is supported by evidence. | Browser-side persistence; server-side saved views. None is supported by evidence. | | Constraints | Export interacts with who is authorized to see which fields, so the permission model matters. Capacity, dependencies and release authority are unknown. | Capacity, dependencies and release authority are unknown. | Gate statuses report the supplied evidence and do not certify safety or compliance. Arithmetic rank and C are preserved despite the HOLD. **Investigation order:** Investigate C2 first (rank 1), then C1 (rank 2). Both remain on HOLD and are subject to remedy fit. **Smallest decision-changing evidence requests** 1. **Both candidates:** the product security lead's completed, dated, scoped Safety and Compliance/privacy review, with the intended use, applicable scope and accountable owner. 2. **Both candidates:** a validated implementation route, so remedy fit can be assessed. 3. **Both candidates:** confirmation of whether the effort estimates include vendor labour and assurance work. This could change the ranking. 4. **C1:** confirmation of the permission model for exports. 5. **Both candidates:** confirmation that EV-101 and EV-102 were observed on v4.2 in EU production. No vendor route is established, so no vendor evidence pack is issued. This analysis grants no release authorization, vendor submission or roadmap commitment (POL-105). ## 5. VERIFICATION AND LIMITS | Check | Result | |---|---| | Source/locator fidelity against original records | PASS (author self-review) | | Identities, business units, dates | PASS. Requester and requested-for are unknown. Business units are attached to submitters as supplied. | | Theme counts and locators | PASS. C1 has 2 raw records, 1 proposition and 1 demand instance. C2 has 1 of each. | | Duplicates and exclusions | PASS. REQ-103 is lineage only. REQ-104 and REQ-105 are excluded. | | Factor transcription | PASS | | Arithmetic table consistency | PASS | | Tie boundary | PASS. T = 5, with the substitution check done. | | Gate/disposition consistency | PASS. Both gates are UNKNOWN, so delivery is HOLD, and the disposition is Further discovery — proposed. | "PASS" here is author self-review only. **Independent arithmetic execution: NOT RUN — displayed calculations are unverified drafts.** **NOT RUN:** - Independent human/source adjudication. - Actual software acceptance testing. - Numerical topic modelling. - Prediction-quality assessment. - Outcome measurement. - Release feasibility. - Extraction precision/recall (no independently annotated propositions were supplied). **Other limits** - The date of each forecast and estimate (EV-103, EV-104) is unknown. - Confidence is a single supplied factor, not a calibrated probability. - Forecasts, estimates and proposed acceptance criteria are not deployment evidence. - This Requests-only corpus is not a full hazard or obligation inventory.