Implement fraud detection with ML-based risk scoring

[Smartdevs17]

Description

Build a fraud detection system that scores transactions in real-time using machine learning models, flagging suspicious patterns for manual review.

Acceptance Criteria

• ML model for transaction risk scoring (velocity, amount, geolocation)
• Real-time scoring pipeline with sub-100ms latency
• Rule engine for configurable fraud rules
• Fraud case management dashboard
• Model retraining pipeline with labeled data
• A/B testing framework for model comparison

Technical Scope

• backend/src/services/fraud-detection.ts
• backend/src/routes/fraud-detection.ts
• Edge: false positive reduction, adversarial inputs, concept drift over time

[Yash-Karakoti]

@Yash-Karakoti has applied to work on this issue as part of the Stellar Wave Program's 5th wave.

i can definitely do this , pls assign !!

ℹ️ Repo Maintainers: To accept this application, review their application or assign @Yash-Karakoti to this issue.