Connect relevant data
Configure the events and signals your process will review.
Loading...
Platform
Organize risk signals, investigations, and decisions with an evidence trail your team can inspect.
Workflows & operations
A risk signal needs context. Fraud workflows help bring the case information, supporting signals, and investigation steps together for review.
Risk scores and AI suggestions are signals, not proof of wrongdoing. They can miss issues or flag legitimate activity.
Platform
Configure the events and signals your process will review.
Use the available context to examine flagged cases.
Keep an accountable history of the outcome and follow-up.
Details and limitations
Define the signals your review process receives and where they originate. Supported checks can use supplied evidence and external scores. Identify which signal is a measured provider result, which is reported by an application and which is a human assessment.
A review record can bring signals, referrals and decisions together so a reviewer can understand why an item needs attention. Cryptographic integrity helps detect changes to recorded evidence; it does not establish whether an unsigned document is authentic from its appearance.
The documented deepfake interface processes supplied signals or scores; it does not independently inspect an uploaded document. Risk values need provenance and context. An unknown or unavailable score should not be converted into a confident fraud accusation or a clearance.
Test a normal fixture, conflicting evidence and a missing upstream signal. Review false-positive handling and the appeal path before using an automated decision. Any outside fraud provider must be configured and accepted against the actual workflow before its results can be relied on.