Invite-only · Master Plan 2 · equity lifecycle
Not a chatbot you start — AI-native agents initiate the entire workflow. Scout, specialist analysts, and the overnight loop observe the book, test living theses, and rank attention before you sit down. Copilot writes into the record; the morning is a huddle you close. Master Plan 2 is the philosophy: knowledge, not documents, is the system of record, and chat is supplemental. Built to future-proof the internal equity investment lifecycle. You still publish, and nothing trades without you.
Access is currently invite-only. Request a walkthrough of the end-to-end loop, or sign in if you already have a seat.
4 tools · budget ok · nothing published
Agents already ran overnight; this session ends in a draft you still own — illustrative, not live market data, and nothing published.
End-to-end AI
Master Plan 2 replaces Read → Excel → Memo → Archive with a living loop: observe, test, update, allocate attention, learn. AI-native agents start every stage of the internal equity investment lifecycle. The analyst still ranks, edits, publishes, and closes. Chat is supplemental — not the organizing paradigm.
Scout, named watchlists, and scheduled screens initiate discovery overnight — ranked against the names you already cover, not a generic idea firehose. Copilot turns news and peers into a shortlist before you open the desk. You tag what enters a pipeline, and what gets ignored.
Company page, dossier, statements, filings, and comps sit together, with Copilot scoped to that ticker. Answers come from tools and the research record — falsifiable claims, not model memory. You stay on the name until the question is actually answered.
Conversation becomes notes and versioned theses inside a pipeline — one claim, staged, attributed, executable later. Knowledge, not the document, is the system of record. You review the draft, edit the argument, and you publish the memo.
Overnight agents flag earnings, thesis drift, and contradictions onto one morning stack. You walk it — confirm, defer, or kill — until the morning is closed. Remaining items are parked, never silently dropped. That is the lifecycle, not an inbox.
Human-in-the-loop
AI-native agents initiate, draft, and test. You triage, edit, publish, and close the morning. Remaining items are parked, never silently dropped. Highest-value decisions stay human — nothing ships, and nothing trades, without you.
The desk
Markets, research, AI, and analysis are rooms on the same future-proof desk — not four products glued together. AI-native agents initiate work in every room; Copilot, Scout, and the morning huddle travel with you so context does not reset.
Named watchlists, charts, statements, and the earnings calendar — plus a ticker-scoped Copilot so company work does not leave the quote. The tape you read at the open is the same book you work all day.
Living notes, thesis pipelines, files, and diligence. Copilot drafts, links, and versions into the system of record — so a session becomes a claim you can kill later, not a transcript you lose.
Scout, coverage analysts, session Copilot, and a facilitated morning — AI-native agents that initiate the workflow. You approve missions and close the huddle; they do not close the day, or the trade, for you.
Supply-chain, geography, and fundamental screens that send names into the same process — from idea to monitoring — instead of a separate analytics island.
Production surfaces
Four live surfaces, already in production — not a chatbot bolted onto a research database. Agents read market data, initiate work, and write into theses, notes, and the morning you still have to close. Master Plan 2 is the frame: observe, test, update, learn.
AI · Research Copilot
Pull statements, search notes, update pipelines, and draft theses without leaving context. Multi-model, tool-driven, bounded autonomy: inspectable trails and capability tiers keep the work accountable. Chat is one interface for querying and updating the record — not the product.
AI · Thesis pipelines
Pipelines hold the equity lifecycle — watch, active research, published memo — so agent output becomes desk work. Versioned, falsifiable claims make drift visible after the print. AI drafts; you keep the record.
AI · Markets & tickers
Named watchlists, fundamentals, and charts for bottom-up work — with Copilot scoped to the name you are studying, so answers stay grounded in that company instead of a generic model.
AI · Analysts, Scout & morning
Personas pressure-test claims with distinct frameworks. Scout initiates weekly discovery against your book. Overnight producers assemble the morning meeting — one beat at a time — until you close it. That is how the lifecycle starts, every day.
Underneath
The same stack as the live portal, built for a living investment world model rather than a better chatbot. Model-agnostic agents, inspectable tools, and a research record that outlives the session — so the internal equity lifecycle can keep compounding.
Model-agnostic by design: Anthropic, OpenAI, and xAI, selected per session — including frontier IDs. Budget ceilings and an inspectable tool trail on every agent turn. T3 is explicit and never auto-picked.
Statements, prices, consensus, and the earnings calendar across domestic and international coverage — the same data the ticker page, Copilot tools, and Scout already use.
When configured, screens and enrichment can use the in-memory engine. Copilot, notes, markets, and the morning do not depend on it.
Notes, pipelines, theses, messages, and audit — held point-in-time so a claim can be replayed after the chat is gone. That is Master Plan 2: institutional memory that compounds across the equity lifecycle, not a transcript you lose.
Invite-only, built on Master Plan 2: agents initiate the equity research workflow; you still close the morning and the memo. Request a walkthrough of the full loop, or sign in if you already have an account.
