Status. Proposed design only. Nothing in this vault is GGA client approval. The GGA client documents are supplied drafts: see GGA-Source-Digest. No application, service or test has been executed; every scenario is Not executed. Delivery fees and dates are open.
INTERNAL — Kedar and the delivery team only. Not for GGA. This whole vault is an internal working vault; see 00-Start-Here.
Internal effort estimates (not for GGA)
Status: history plus a re-estimation checklist. The figures below estimated the 1.3 private-exchange scope. They do not cover the GGA scope and must not be reused for it. No GGA estimate exists; none is invented here. Planning judgement only, not fees, not a quote, not a staffing commitment.
1.3 figures (history)
| Option | Total engineer-days (incl. risk allowance) | Two-person planning window |
|---|---|---|
| Optimized | 36–48 | 4–5 weeks |
| Balanced | 65–90 (base 54–75 + 20% contingency) | 7–9 weeks |
| Comprehensive | 110–150 (design/QA support extra) | 11–15 weeks |
Balanced build-up then: Optimized foundation 30–40; contributor, approver and viewer roles with version-aware approvals 8–12; outbox, worker, email and retries 8–12; pending and overdue view 3–5; extra permission, worker and regression evidence and UAT corrections 5–6. Confidence was low to moderate: no code or client infrastructure was inspected. No hourly rate was recorded. The illustrative stack then was about $75–95 a month for Optimized without AI; Balanced needed a worker and queue budget that was never verified. Assumptions then: two dedicated experienced engineers; engineer-led QA; one consolidated testing round; no migration, SSO, external integration or penetration test. Sources: Technical-Review-Original and Tier-Technical-Verdict.
What the GGA scope adds (to be sized in Stage 1)
| Area | New or changed work | Blueprint |
|---|---|---|
| Public site and projection | Approved-only public API, caching rules, withdrawal invalidation | pp. 5, 14, 17 |
| Profiles and freshness | Structured profile, versions, attestation, evidence, reminders, request-time eligibility | pp. 4, 7 |
| Partners and review | Partner approval, intake, three information layers, operations review queue | pp. 5, 13 |
| Introduction and consent | Pitch, scoped fields, consent and grant records, one-time release, withdrawal | pp. 6, 15 |
| Files | Private storage, type and content checks, scanning, grant-checked download | pp. 13, 17 |
| Events and sessions | Occurrences, registration, capacity, waitlist, reminders, calendar files | p. 9 |
| Outcomes and contribution | Outcome records, opt-in, finance console, invoice, reconciliation, 12-month terms | pp. 8, 15 |
| Jobs and exceptions | Job catalogue, retry schedule, exception queue, retry action | p. 16 |
| Security and operations | MFA, row-security testing with the real role, export and deletion workflows, monthly restore, performance test | p. 17 |
| Delivery | OpenAPI with examples, field dictionary, configuration register, handover package, acceptance tests (Blueprint A1 to A12, vault GA01 to GA12) | pp. 15, 18, 19 |
What carries over from 1.3 effort: foundation and CI, outbox and worker, isolation harness, version-aware approvals, and the evidence discipline. Do not copy the old numbers into any GGA-facing material without the owner’s decision. The “services DG-funded” and “illustrative stack” lines in the 1.3 estimate must be re-derived against the providers GGA chooses.