Product · Applied AI Live
Ten AI analysts, one decision, a full audit trail.
A spot-only accumulation engine. Independent AI analysts each read live market data, a consensus step reconciles them, and every decision is recorded with the reasoning that produced it.
Python · React · TimescaleDB · FastAPI
What it does.
Consensus, not a single oracle
Independent analysts assess the same market data from different angles. A disagreement is information, and the consensus step is where it gets resolved rather than averaged away.
Spot only, long only
No leverage, no shorting, no derivatives. The engine accumulates assets it can hold outright, which bounds the downside to the market's.
Every decision is evidenced
Each trade carries the analyst inputs and the reasoning that led to it. You can reconstruct why a position exists months later.
Continuous evaluation
Candidate strategies run in parallel against the live book as paper lanes, so a change earns its place on forward evidence rather than a backtest.
How it works.
Market data streams into a TimescaleDB time-series store.
Each analyst evaluates the current state independently.
A consensus step reconciles the analysts into one decision.
Risk limits are applied before anything is submitted.
The decision, its inputs and its reasoning are written to the run store.
A React dashboard exposes positions, decisions and history.
Worth knowing
Not investment advice
This is engineering, not financial advice, and nothing here is an offer or a recommendation. Past behaviour of any strategy does not predict future results.
Where it runs
Self-hosted, against your own exchange API keys. Keys stay in your environment.
Alpha Engine in your environment.
It is running now — open it, or ask about running it in your own environment.