Architecture & roadmap
What's built, what's planned, and in what order.
One founder plus one engineering hire, a 40-week plan, and a strict rule: the wedge works without any of the expansion modules. Each module is labeled with its build status. Nothing unbuilt is described as though it exists.
Core architecture — built
The pipeline that exists today.
Sanitize → triage → synthesize → guardrail → execute → learn. Unstructured intent in; typed, dependency-ordered, tool-executed actions out. Per-workspace decision memory on pgvector with HNSW indexing, sub-50ms retrieval. MCP-native protocol interface. Live execution in Linear, Slack, Notion, and Gmail. The full mechanism is on the Product page →
Average model cost per directive processed: $0.0004. That number is measured, and it's why the unit economics have room to work.
Expansion modules
Seven modules on the roadmap. Four are far enough along to name.
Internal codenames included because they'll come up in diligence — with a plain-English name and an honest status on each.
Decision memory — internally “VAULT”
The per-workspace store of every card, outcome, and rationale. Live today in the core pipeline; the expansion adds deeper retention analytics and cross-decision pattern detection.
Org rollups & dependency detection — internally “ENSEMBLE”
Org-wide rollup briefings, approval policies, and cross-team dependency detection. This is the upmarket move into Ops/PMO once decision-memory volume is real.
Pre-execution simulation — internally “SIMULACRA”
A dry-run of a directive's full execution plan before anything is sent — what will be created, who gets notified, in what order.
Proactive surfacing — internally “PRECOG”
Flags decisions that appear to be stalling — no owner confirmed, deadline slipping — before anyone has to ask.
Three further modules are in planning and deliberately unnamed here. Codenames for unbuilt software read as vaporware in diligence; these four get named because they're sequenced and scoped.
Roadmap
The next 40 weeks.
-
Weeks 1–8 — Prove the core loop
Ship the core loop into five to eight design partners matching the ICP (ops/eng leads, 20–150 people, Linear + Slack). Instrument decision-memory usage and retention, not just directives processed. Exit: a one-page proof doc with measured retention, time-to-first-value, and seat-expansion signal.
-
Weeks 9–16 — Harden trust
Approval policies, deeper guardrail configuration, and the decision-memory expansion (VAULT). The first engineering hire starts here, funded by this raise.
-
Weeks 17–28 — Move upmarket
Org-wide rollup briefings and cross-team dependency detection (ENSEMBLE), gated on real decision-memory volume from Phase 1.
-
Weeks 29–40 — Anticipate
Pre-execution simulation (SIMULACRA) and proactive surfacing (PRECOG). In parallel: the seed round, sized to the evidence collected in Phase 1.
Use of funds
The $50,000, split 40 / 25 / 20 / 15.
Planned allocation. It changes only if design-partner data says it should.
Engineering and Product
The first engineering hire and founder time on the core loop in design-partner workflows — fixing what breaks in real usage and instrumenting the memory and retention numbers the next round depends on.
Go to Market
Recruiting the five to eight design partners, onboarding them into real workflows, and turning their usage into the measured proof doc that replaces every modeled claim on this site.
Infrastructure and AI
Hosting, decision-memory stores, and model spend. At $0.0004 per directive, the AI bucket is mostly headroom — this buys reliability and margin, not scale.
Runway buffer
Legal, accounting, and the surprises between week one and week eight. A buffer so the proof phase isn't one invoice away from slipping.
Honesty check
What's de-risked on the technical side.
Already true
- End-to-end execution in four tools: Linear, Slack, Notion, Gmail.
- Decision memory live with sub-50ms retrieval.
- Guardrail stage in front of every execution path.
- MCP-native interface — new tools are protocol connections.
Still to prove
- Decision-memory retention compounding — the moat thesis, measured in Phase 1.
- Seat-by-seat expansion inside a real org.
- Reliability at design-partner scale, not founder-demo scale.
- Enterprise requirements (SOC 2 and friends) — on the roadmap, not in hand.
Technical diligence welcome.
Schema, policy engine, protocol layer — ask for the internals.