Use cases
Fraud prevention
Put several checks in one flow, each aimed at a different way of cheating.
The job
Fraud against an identity check takes several forms, and one check cannot answer them all. A forged document needs one test, a replayed photo another, a generated face a third, and a person who keeps coming back a fourth. This flow lines them up in one session.
What each check is aimed at
| The attempt | The check | Read more |
|---|---|---|
| A forged, copied or edited document | Document anti-fraud with 20+ checks | Document authenticity |
| Someone else's document with a different face | Face match between the selfie and the document portrait | Face verification |
| A photo, a screen or a mask in front of the camera | Active liveness, a movement challenge through two ovals | Liveness |
| A generated or swapped face | Deepfake and face-swap detection | Deepfake detection |
| The same person again and again, or a listed person | Repeat-user search and list search | Face search |
The flow
- The document. The user photographs an identity document. It is read, and the anti-fraud checks examine the image.
- The face. The selfie is taken and matched to the document portrait.
- The live person. The user moves the face through two ovals. Deepfake and face-swap detection runs inside the selfie and liveness steps.
- The history. The face is searched among earlier verified faces and checked against your blocklists.
- The review. Your rules read the results. Cases you want to look at go to a manual decision in the dashboard.
Scenario setup
The document, face, liveness and search checks are steps and options that you choose in the scenario editor. Use a lighter scenario for low-risk moments and a fuller one for the moments where the cost of fraud is high.
What your system receives
A session status and a list of results for each step, so your rules can weigh the layers separately. A session can also be flagged as suspicious. See session statuses.
Next
See the layers on sample data in the interactive tour, or talk to us about the fraud patterns you face.
Questions, answered plainly
Which checks does the flow combine?
Document anti-fraud with 20+ checks, face match, active liveness, deepfake and face-swap detection, and repeat-user and list search.
Does a failed check reject the user automatically?
The result is added to the session. Your rules decide what happens next, and a reviewer can make a manual decision in the dashboard.
What does the flow give my team?
Several checks in a row, results that your rules can read, and a place in the dashboard to review the cases it wants to see.