What we collect, which is very little
Data minimisation is not a policy position here, it is a design constraint that shows up in the code. Where we can avoid holding something, we do.
This website
No account is required to read anything here and there is nothing to log into. We do not sell data and we do not run third-party advertising trackers.
The build tracker and the changelog read repository data server side. That involves no information about you.
Exchange API keys, which is the important one
Keys for the trading engine are encrypted and stored on your own machine. They are not uploaded, not escrowed, and not recoverable by us. That also means we cannot help you recover them if you lose them, which is the cost of the guarantee being real rather than a marketing line.
This is a claim that can be verified in code rather than believed, and verifying it is a standing part of the review every unit of that system goes through before it is called finished.
What the chain records about the trading engine
The Signal Ledger records the asset, the direction, the confidence score with its per-engine breakdown, the model version, a timestamp, and a hash commitment of the inputs. It never records your exchange identity, your API key, your position size, your fill price, your profit or loss, or your balance. A public record of decisions is useful. A public record of your positions would be a liability with your name on it.
Community systems
The Discord and Telegram systems process the identifiers those platforms give us in order to attribute referrals and score engagement. Full terms are published before that system opens, and they include our right to claw back rewards obtained through fake accounts or coordinated gaming.
Subscriptions
Billing runs on our own infrastructure and is one of the few things we openly centralize. Detailed terms are published before checkout opens. Card details are handled by the payment processor rather than stored by us.