Decline the fraud, keep the customer, and be able to explain both.
A synchronous verdict inside the authorization path, carrying the score, the rules that fired, the model version and an audit reference. The pack is built, which makes it the fastest route into a first pilot on your data.
// response · 87ms
{
"action": "review",
"score": 0.91,
"severity_band": "high",
"reason": "velocity + geo mismatch",
"evaluators": [
{"rule": "cnp_velocity", "hit": true},
{"model": "txnCBM@v14", "score": 0.91}
],
"audit_ref": "dec_EXAMPLE…",
"latency_ms": 87
}
One question, answered the same way every time.
Card, account, device, velocity, geography, merchant and behavioral history, on point-in-time-correct features.
Action, score, severity band, the rules that fired, the model version, and an audit reference.
Fraud strategy, fraud operations and financial crime, in Command Center and Workbench.
The verdict stays lean. The investigation runs beside it.
A hard timeout budget and a fallback verdict when it is breached. Fail-open or fail-closed is your policy, set in configuration.
Related cases, device history, dispute context and policy evidence assemble before an analyst opens the case.
Shadow first, enforce second, on your own history.
Your history, replayed
Policies simulated against your own decision history, so you can see what would have changed before anything changes.
Live traffic, acting on nothing
Nyx scores beside your existing checks. You compare its calls with your team's, on your traffic.
Narrowly, with your rollback
Turn it on for a scoped segment, with your fallback verdict and your rollback, on evidence from your pilot.
Deployment is your choice in every case: managed by us, in your cloud, or in your data center. Compare the deployment models →