Skip to content

Fraud detection

Traceveil

Spots fraud from how people behave: dodgy transactions, bot rings and exam cheating, with answers in under 50 milliseconds.

Traceveil, sketched

Context

The problem

Fraud signals are scattered across payments, sessions and behaviour, and you only get a few milliseconds to decide.

Approach

How I built it

  1. LSTM and autoencoder models score how unusual the behaviour looks across different kinds of signals.

  2. The scoring is built to be fast, with a target of under 50ms per decision.

  3. Monitoring was part of the plan from the start, for payment risk, bots and exam integrity.

Result

What came out of it

  • Decisions in under 50ms

  • One system for three kinds of fraud

  • Monitoring built in