Risk controls section

Fraud & AML Engine

See how product services send normalized risk events, how a provider adapter can call Sardine-like fraud, AML, and KYT services, and how decisions come back before money movement or regulated card activity.

4 decision outcomes 3 integration flows 1 provider-neutral boundary

Interactive flow lab

Step through fraud and AML decisions

Move from checkout risk to KYT and AML screening, then into the Sardine-style provider adapter.

High-level sequence

Decide a transaction: allow, step-up, review, or block

Every money movement passes one risk decision with real product outcomes, not a silent filter.

Customer Product service Risk engine Fraud provider KYT provider

1

Customer→Product servicerequest

Initiate a transaction — send, spend, or cash out.
2

Product service→Risk enginerequest

Request a decision with full context.
3

Risk engine→Fraud provideroff-platform

Device, behaviour, and velocity signals.
4

Risk engine→KYT provideroff-platform

Wallet / counterparty and sanctions screening.
5

Risk engine⇢Product serviceresponse

Decision — allow, step-up, review, or block.
6

Product service→Customerevent

Proceed, request step-up (OTP / 3DS), or hold.
7

Risk engine→Ops back-officeevent

Open a case on review / block, with an audit trail.

Interface map

Find the risk boundary

Filter the fraud engine calls by product event, AML control, provider adapter, and operations.

Concept notes

The risk ideas worth remembering

Questions

Check your understanding

Pick an answer and the page will explain the reasoning.