01The two-clock problem
Real-time fraud detection has two clocks running. The system clock: decisions in under 200ms. The labeling clock: ground truth (chargebacks, confirmed fraud) arrives days to weeks later. You design around both.
A reference architecture for real-time fraud detection — features, models, decision policy, and the false-positive trade-offs that make or break customer experience.
Binate AI
July 21, 2025

On this page
Real-time fraud detection has two clocks running. The system clock: decisions in under 200ms. The labeling clock: ground truth (chargebacks, confirmed fraud) arrives days to weeks later. You design around both.
Every transaction lands on a topic. Schema-versioned. Replayable.
Aggregations over windows (1m, 1h, 24h, 30d) maintained by Flink or a feature store. Latency: <30ms.
Gradient-boosted tree or small NN. Latency: <50ms. Logs every input and score.
Score + business rules → approve / step-up / decline. Tunable thresholds per merchant or card type.
Chargebacks, customer disputes, and manual reviews flow back as labels. Retrain weekly, evaluate against a frozen test set.
Fraud detection is a UX trade-off, not just a math problem. Every false positive is a real customer locked out of their card. Every miss is fraud loss.
0.2–0.5%
Target false positive rate
70–90%
Detection rate (per segment)
<200ms
End-to-end decision
Approve / decline is a false binary. A "step-up" path — challenge with 3DS, SMS, or biometric — turns false positives into recoverable friction.
Quick Quiz
We build real-time fraud detection for fintech and payments.
See our fintech workFraud detection is a real-time system, a feedback loop, and a UX trade-off — all at once. The teams that get all three right ship; the others stall at 80% accuracy and a churn problem.
A real-time fraud detection system sends each transaction event through Kafka to a feature service that maintains windowed aggregations in under 30ms, then to a model such as a gradient-boosted tree that scores it in under 50ms. A decision policy combines the score with business rules to approve, step up, or decline, and chargebacks and disputes flow back as labels for retraining.
Our experts are ready to power your AI journey.