reference
MemberIntel Roadmap — where we actually are
The single source of truth for where MemberIntel stands: the two numbering schemes (product Versions vs calendar Phases) untangled, the full shipped surface as an expanded V1, and the next real milestone (V1.5). Status for ADR-backed rows auto-refreshes daily.
Last reconciled: 2026-07-01. Status for ADR-backed rows refreshes automatically in the daily pipeline; the rest is reconciled by hand against the PR log + standups (see “How to update” at the bottom).
Read this first: two numbering schemes, not one
Most of the “where are we?” confusion comes from mixing two different axes that both use small numbers:
| Axis | Values | Means |
|---|---|---|
| Product Version | V1 / V1.5 / V2 | Capabilities. V1 = advisor + dashboard + two-tier brain (read-only). V1.5 = AI write-actions on the customer’s MemberPress site (via the MP MCP) + a “Set up FOR ME” onboarding wizard. V2 = BuddyBoss. |
| Team-ramp Phase | Phase 1–5 | Calendar. Phase 1 = May scoping … Phase 5 = launch. A staffing/timeline plan (phased-plan-v2), now drifted. |
“Between phase 1.5 and phase 2” glues a Version label (1.5) to a Phase label (2) — it isn’t a real position. Use Version for “what can the product do” and Phase only as historical calendar context.
You are here: an expanded V1
We’ve shipped a V1 that’s far richer than the original spec defined — multi-site portfolio, a full admin suite, session continuity, evals, cohort intelligence, profile-aware retrieval, and more. That breadth is why it can feel past 1.5.
But V1.5 as a milestone has a specific bar — the AI executing approved changes on a customer’s live MemberPress site, plus the “Set up FOR ME” wizard — and that bar is unstarted. So the honest read:
We’re deep in an expanded V1, with the brain/moat as the active frontier. The next milestone (V1.5 = write-actions + wizard) has not begun. We are not “between 1.5 and 2.”
Everything shipped to date (by capability domain)
The real breadth — not squeezed into three version buckets. shipped = live in production; in progress = actively being built; held = built-spec, deliberately shelved; not started = next milestones.
Shipped — V1 core + expansions
| Domain | Status | Evidence |
|---|---|---|
| Chat/advisor (SSE streaming, tools, tool-result snapshots, ±1 feedback, escalate-to-human) | shipped | ADR-0023; PRs #99–123, #228–230 |
| Per-customer brain (SOUL/BIBLE/HEARTBEAT/MEMORY) | shipped | ADR-0015/0016/0018 |
| Session continuity / wrap-up logs | shipped | ADR-0019 |
| Multi-site / portfolio chat | shipped | PRs #68, #73–85 |
| Retrieval / RAG (Voyage 3-lite, pgvector, tenant isolation, Hive Mind seed) | shipped | ADR-0006/0008 |
| WP-connect plugin + data sync (one-click connect, v1.6.1) | shipped | ADR-0017; PRs #164–173 |
| Dashboard / KPIs / trends / portfolio rollups | shipped | PRs #50, #56–78 |
| Auth + account lifecycle (Google OAuth, JWT versioning) | shipped | ADR-0024 |
| Admin operator suite (SPA, user mgmt, brain-vault editor, audit log) | shipped | PRs #123, #127–146 |
| Evals / regression-replay | shipped | PRs #106, #111, #182 |
Entitlement / quota (check_and_consume, upgrade nudges) | shipped | ADR-0001; PRs #1, #54 |
| Model routing — tier-pinned Anthropic (Haiku Free / Sonnet Pro) | shipped | ADR-0025; PR #160 |
Multi-provider routing (Gemini via llm.call seam) | seam shipped; rollout deferred (not needed now) | ADR-0020 |
| Infra / CI / observability (Cloud Run, GCP, WIF, BigQuery, OTel) | shipped | ADR-0003/0004/0009–0013/0022 |
| Marketing site (membersintel.com — Astro, 5 pages) | shipped & live; design/copy sign-off pending | caseproof/memberintel/site; innovations #113/#125/#126/#127 |
Active frontier — in progress
| Domain | Status | Evidence |
|---|---|---|
| Global-brain content pipeline (playbooks, VOICE.md, ingest spine) | shipped | ADR-0026; innovations #11/#154/#12 |
| Cross-pollination review/promote (candidate staging → embed-on-approve) | in progress | PRs #179–190, #222, #239 |
| Cohort benchmarks + profile-aware retrieval | shipped | ADR-0028; PRs #200–223 |
AI-Interviewer primary_research brain feed | staged; prod-gated on counsel | innovations #156/#147 |
Deliberately held (built-spec, shelved launch-first)
| Domain | Status | Evidence |
|---|---|---|
| Subscription-events sync | held | spec #88 |
Read-only query_sql surface | held | spec + 14-task plan |
Not started — next milestones
| Domain | Status | Evidence |
|---|---|---|
| V1.5 — AI write-actions agent (via MP MCP) | not started | v15-spec §1 |
| V1.5 — “Set up FOR ME” greenfield wizard | not started | v15-spec §2 |
| V2 — BuddyBoss integration + platform-tagging | not started / planning | v2-spec; innovations #165/#164 |
Known gaps (real state, not blockers)
| Domain | Status | Evidence |
|---|---|---|
| Billing execution (Stripe) | stub only | PR #132 — entitlement real, payment not wired |
Version milestones (the “what’s the next lift” view)
- V1 (original spec): advisor + dashboard + two-tier brain, read-only → shipped.
- V1 as-built (expanded): + multi-site portfolio, admin suite, session continuity, evals, cohort intelligence, profile-aware retrieval, global-brain pipeline → shipped + active frontier.
- V1.5 (next milestone): write-actions agent + “Set up FOR ME” wizard → not started.
- V2: BuddyBoss + platform-tagging → not started / planning.
Brain data sources
The brain (the moat) is fed by: the WP-connect plugin (members/transactions/subscriptions/memberships), Stripe (payment history), the Hive Mind seed (MP product knowledge), cross-pollination (anonymized patterns, V2+ for full promotion), and the AI Interviewer primary_research feed — a sibling product (caseproof/interviewer) whose analysis output feeds the brain via an adapter that’s built + verified on staging, with production publish gated on counsel sign-off (innovations #156).
Work in flight / what’s next
Tracked as [MemberIntel]-prefixed issues (currently in caseproof/innovations; the tracker is mid-reorg, so some may move — identify them by the title prefix, not the repo):
- Brain content: author first 50 playbooks (#11), cross-pollination review process (#12), staged ingest spine (#154).
- Marketing site polish: IA + copy (#125), design (#126), build (#113), launch cutover (#127).
- Launch prep: cost-control circuit-breaker (#22), sales/CS enablement (#21), PR agency (#20).
- Ahead: V2 integrations planning (#165), WordPress Dashboard build (#164).
Calendar reality
phased-plan-v2 projected a May–Oct ramp (Phase 1 scoping → Phase 5 launch). Those dates are not being re-forecast — some are already past, others will push out. Treat that page as the historical staffing/calendar plan; treat this page as current status. We track by sequence and status, not month targets.
How to update this page
- Auto (daily): status cells wrapped in
<!--rs adr=NNNN-->and the “last reconciled” date refresh automatically from ADRimplementation_statusvia.github/scripts/sync-roadmap-status.mjs. Don’t hand-edit those cells. - By hand: every other row (no marker), all prose, the inventory membership, and the work-in-flight list. Re-reconcile against the product PR log + standups periodically — search by path and treat membersintel.com as ground truth (the audit method that seeded this page).