Product

Customer Risk Profiling

One authoritative risk state per customer, updated in real time. Every transition is triggered by an event, approved by someone accountable, and written to an immutable trail.

The problem

Risk signals scattered across systems produce contradictions

Risk signals are scattered: a screening hit lives in one system, a monitoring alert in another, KYC status in a third. Without a single authoritative state, two parts of the same fintech can hold contradictory views of the same customer — and no one can reconstruct why a customer was treated the way they were.

Without a unified risk state

  • Screening result sits in the vendor dashboard, not connected to payment decisions

  • Monitoring alert raises a flag, but the payment handler doesn't know

  • KYC tier stored in onboarding system, not enforced at transaction time

  • Two engineers query different systems and get different risk answers for the same customer

  • Evidence reconstruction takes days when a regulator asks why a transaction processed

With Fintegrity's unified risk state

  • One state per customer — read by the Decision API before every transaction

  • Screening hits, monitoring alerts and KYC events all converge into the same state machine

  • Every state transition recorded with its trigger, approver (where required) and evidence

  • Any query returns the same answer — the authoritative current state

  • Full history of how a customer's risk evolved reconstructable on demand

The risk state machine

Six states. Event-driven transitions. Every move evidenced.

Fintegrity maintains a single risk state per customer, driven by a compliance state machine. The state moves only through defined, event-driven transitions — and each transition records what triggered it, who approved it (where approval is required), and what evidence was used.

States in detail

What each state means for a transaction

The risk state is read by the Decision API before every transaction. It is not a report you look at after the fact — it is enforced at the moment of every call.

KYC_PENDING

Onboarding started

Verification has been initiated but is not yet complete. Transactions may be limited to a minimum tier until verification passes.

KYC_OK

Verified and in good standing

The customer is fully verified. Transactions proceed through full rule evaluation with no additional restrictions from risk state.

KYC_REFRESH_DUE

Verification due for refresh

Verification has expired or reached its scheduled refresh date. Elevated monitoring applies until refresh completes.

RISK_REVIEW_REQUIRED

Flagged for compliance review

Triggered by a screening hit, monitoring alert, or manual escalation. A case is open; an analyst is investigating.

RESTRICTED

Limited access pending resolution

High-risk case outcome. Access is restricted while the investigation or remediation is in progress.

BLOCKED

Barred from transacting

All transactions return BLOCKED before any rules run. No amount or type of transaction can proceed while this state is active.

Transition audit

Every state change is an evidence event

A manual override requires a reason, the approver's identity, and supporting evidence — no silent state changes. Automatic transitions (from screening hits or monitoring alerts) record the triggering event directly in the transition record. The result is a full chronological history of a customer's risk lifecycle, reconstructable on demand.

Trigger

What caused the change

Every transition records the triggering event — a screening match, a monitoring alert, a case outcome, KYC expiry, or a manual action.

Approval

Who authorised it

State changes that require manual action are attributed to a named analyst. Automated transitions are attributed to the system event that triggered them.

Evidence

What supported the decision

Supporting evidence — case reference, screening result, rule that fired — is linked to each transition and retained in the immutable audit ledger.

Enforcement in the decision layer

The profile is enforced, not just recorded

Because the risk state is read by the Decision API before money moves, the profile isn't a report you look at after the fact — it's enforced at the moment of every transaction.

A BLOCKED customer gets a hard stop before any rules run. It doesn't matter what device they use, what channel they try, or how small the transaction amount is — the state check happens first, and it's deterministic.

And because every transition is evidenced, the full history of how a customer's risk evolved is reconstructable on demand — for regulators, auditors, or your own compliance team.

See how Fintegrity evaluates transactions in real time

We'll walk through the risk state machine live — from a screening hit triggering a state change, to a case outcome resolving it, to the evidence trail that makes every decision defensible.