AI-Powered Anomaly Detection & Fraud Prevention Engine #406
Description
Description:
Overview
Build an intelligent fraud detection and anomaly prevention system that uses machine learning to identify suspicious transactions, unusual spending patterns, and potential security threats in real-time.
Requirements
Backend - Models
Create AnomalyRule model: name, type (threshold, pattern, velocity, geo, behavioral), conditions JSON, severity (low, medium, high, critical), action (alert, block, review), isActive, createdBy
Create AnomalyEvent model: userId, transactionId, ruleId, type, score (0-100), details, status (pending, confirmed_fraud, false_positive, resolved), reviewedBy, reviewedAt
Create UserBehaviorProfile model: userId, avgDailySpend, avgTransactionSize, typicalCategories[], typicalMerchants[], activeHours[], typicalLocations[], deviceFingerprints[], lastUpdated
Create RiskScore model: userId, overallScore, factors[], calculatedAt, trend (increasing, stable, decreasing)
Create BlockedEntity model: type (merchant, ip, device, card), value, reason, expiresAt, addedBy