Platform
One workspace for
the whole discipline.
The engine grades your decisions — the workspace is where you make them. Everything you research, declare, watch, and trade lives in one private workspace, wired to the same grading engine and shared only when you say so.
The founding build ships the core loop first; the rest is built in the open with the founding cohort.
Your workspace
Every artifact in one place — and everything gets graded.
Journal
Every trade, pass, and reason captured as you decide — not reconstructed after. The journal is the raw material the engine grades you on.
Watchlists
The names you track, wired into the machine: setups on your list are logged, simulated, and studied whether you act or not.
Studies
One-shot questions answered from the record with sample sizes — and saved, so a settled question stays settled.
Theses & plans
Falsifiable claims with immutable evidence, each carrying its management plan. Published theses keep a steppable version history — revisions are visible, never silent.
Portfolios
Paper and live books side by side, graded by the same engine — with every real fill audited against its simulated twin.
Risk policies
Declare your own rules — position size, exposure, loss limits. The platform grades you against them, and if you choose, enforces them as a hard block you opt into yourself.
Files, wiki & meetings
Datasets and documents live with the work, the team wiki keeps the thinking, and meetings happen inside the workspace — with transcripts the agent can cite.
Winners and losers, over time
Every thesis cohort is charted as it ages, so the record shows drift the moment it starts.
Your agent
An AI that remembers your whole book.
The agent isn't a chatbox bolted on — it lives in your workspace, keeps memory of your theses, your record, and your open questions, and carries context from conversation to conversation.
It co-authors theses with you, writes and backtests the indicators and screens you invent, and answers from a trading knowledge base the platform maintains — quoting the evidence, including when the evidence says no.
- Collaborative authoring — declare theses in conversation; the agent drafts, you decide.
- Persistent memory — your agent knows what you asked last month and what the record said.
- Grounded answers — studies run on the platform's record, sample sizes quoted, "nothing there" included.
- Everywhere you work — pop the agent into its own window, pick the model it runs on, or show it what you mean on the live page.
- Bring your own AI — plug in your own AI vendor keys: same tools, same guardrails, your account.
Every lens
Every kind of analysis. One discipline.
However you read a market, the workflow is the same: state the claim, attach the evidence, validate it forward, get graded. No lens is second-class — and no lens is exempt.
- Fundamental — company and business health, and value.
- Technical — price, volume, trends, and chart behavior.
- Quantitative — mathematical, statistical, and systematic analysis.
- Sentiment — investor, news, social, and analyst sentiment.
- Macro — economy, rates, inflation, currencies, commodities.
- Market — overall market, sectors, industries, breadth, and regimes.
- Options — volatility, positioning, Greeks, skew, and derivatives.
- Risk — downside, volatility, correlations, exposure, scenarios.
- Event & catalyst — earnings, news, M&A, regulatory events, product launches.
- Portfolio — allocation, diversification, exposure, and portfolio-level effects.
And the qualitative layer runs through all of it — narratives, catalysts, management quality, your reasoning — captured as first-class evidence when you declare, so even the soft calls face the scoreboard. Missing a lens? The agent builds the tooling in conversation.
Protocols
Custom power, sealed and watched.
Protocols are the platform's unit of custom capability: agent-written indicators, screens, scoring systems, and studies, scheduled data pulls from your own sources, and sealed programs that run entirely inside a hard sandbox — backtested with a time holdout, so an overfit idea indicts itself before it touches a decision.
Every Protocol is versioned, comparable run to run, and monitored against public vulnerability feeds — with an optional dial that pauses a Protocol automatically if a critical flaw is disclosed. Your data joins the same graded workflows.
- Agent-built Protocols — ask for the tool; it's written, sandboxed, and backtested before it can influence anything.
- Bring your own data — your feeds, subscriptions, and APIs, connected under your own keys and kept in your workspace.
- Datasets that persist — studies and datasets live in your workspace, reusable and citable in every thesis that follows.
- As fast — or as deliberate — as you need — studies, Protocols, and refreshes run at the pace you set, from tick-by-tick to daily with one-minute resolution as the default; the platform provisions the performance to match.
Execution & alerts
Signals reach you. Orders never leave you.
Signals & alerts
When a live signal fires it reaches you by priority — critical alerts interrupt, everything else batches into digests you control.
Broker connection
Connect your own brokerage account for the live loop: verified fills, audited grades, and a track record the platform generates — not one you type in.
Your authority on every order
The platform never trades on its own initiative. Every order that reaches the market runs on authority you granted — your click, or a plan you armed yourself. That guardrail is permanent.
Automation on your terms
Entry and management strategy plans run as order-management automation — the engine executes every plan deterministically, no hindsight, no exceptions — or you keep it to signals and alerts and decide each one yourself.
Crypto, first-class
Crypto joins stocks, options, and futures — connect your own US venue account; every order on your authority, every fill graded.
In your pocket
Thesis installs as an app on iPhone and Android — no app store, auto-updating, push alerts to your lock screen. See mobile for what travels with you.
Voice, video & time
Rooms that talk, calendars that keep up.
Calls in every room
Voice and video native to communities and group chats — with recordings and transcripts that become searchable artifacts your agent can read back to you.
Events & masterclasses
Community events, ticketed sessions, and time-boxed masterclasses that start, run, and archive — hosted where the artifacts already live.
Calendars that sync
Workspace calendars for market events, thesis reviews, and community sessions — with two-way sync to the calendar you already use.
Reach
Publish once. Land everywhere.
Published theses, track records, and community moments can crosspost to your social channels — polished by the agent, previewed by you, sent under your name.
Daily or weekly digests keep members current without living in the feed; notification priorities mean the only interruptions are the ones you chose.
- Everywhere you are — the full experience in the browser and as an installable app on iPhone, Android, and desktop. See mobile.
- Critical alerts that land — time-critical signals reach you on the channel that actually interrupts, not just another push.
- Your data, exportable — your artifacts are yours: full exports, no lock-in theater.
Teams & institutions
The same discipline, at desk scale.
Organizations get their own isolated workspace: shared theses and research, team structure with typed seats, and admin control over what members may join and publish in the public layer.
Teams
A shared research desk — org workspace, member management, and the whole grading discipline applied to the desk's book, not just one trader's.
Enterprise
Governance included: directory sync, nested team structure with inherited visibility, admin oversight, and isolation with hard guarantees.
Official & institutional
Verified brands operate Official communities, and institutions can exchange research and data with the platform on negotiated terms — a data relationship, not an ad relationship.
Organization plans open after the founding cohort — early conversations shape them. The full picture is on the Team and Enterprise & Platform pages; register early interest or start with Security & compliance.
The workspace is built in the open.
Founding members get the core loop first and shape everything that follows. Make your case.