Healthcare Insurance Claims Fraud Detection System

Build an advanced machine learning model using TensorFlow and pandas to detect insurance claims fraud in healthcare. Develop a multi-layer neural network that can analyze claim patterns, flag suspicious transactions, and provide a risk score with explainable AI techniques. The system must handle complex feature engineering, support real-time inference, and generate detailed compliance reports. Implement a modular architecture that allows for continuous model retraining and performance tracking.