Evidence, or an honest no
Ask about a case. Get the answer from what happened to cases like it.
Will this customer churn? Which contract keeps them? Which 50 accounts first? Send a record of past outcomes and your questions. Datagoat answers each case with a chance, the columns that moved it, and a signed Verdict. When the record cannot support an answer, it says so, and the refusal is free.
4.48× lift on held-out rows, synthetic churn sample
1,000 fits a month, free
$20 per million answered cases
Four question types. One call.
| Type | Answers |
|---|---|
yesno | the chance the outcome happens |
score | that chance as a level: unlikely, possible, likely, very likely |
choice | the best option: the likeliest to give an outcome you want, the least likely to give one you avoid |
rank | cases in order of that chance |
Your data, in the shape it is in.
A table with one row per case, an event log, a weekly series, a panel of monthly periods, a stream of sensor readings, a log of agent runs. Name the shape and the engine reads the record into rows per case before it fits. Shapes.
The bar never moves.
- Deterministic. The same record and the same question give the same answer, every time.
- Glass box. Every case lists up to four columns that moved its chance, and which way.
- Refuses rather than guesses. No pattern that holds on held-out rows means no answer, and no charge.
- Signed. Anyone can check an answer was not altered.
- Nothing kept that you did not ask us to keep. Inline data is deleted when the call ends; stored datasets 24 hours after last use. Your data.
Connect
MCP https://api.datagoat.io/mcp REST POST https://api.datagoat.io/v1/ask Python pip install datagoat TypeScript npm install @datagoat/sdk