Major project
Banking & Fintech
Real-time payment fraud detection with explanations
Catch fraudulent payments as they happen, with a reason an investigator can check.
The brief
A digital payments company must stop fraud without blocking genuine customers. Build a fraud-detection system for card and mobile-money transactions that scores each payment in real time, explains its score, and lets investigators review flagged cases. Good looks like: high fraud catch at a low false-alarm rate, a cost-based threshold, explanations, and a working scoring service.
What you will build
- Explore fraud patterns and the class imbalance
- Engineer transaction and behaviour features
- Train models and handle imbalance properly
- Choose a threshold from the cost of fraud vs blocked customers
- Serve the model as an API with explanations
- Build an investigator review screen and write the report
Free, open data to build it with
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.