AI-powered heart disease prediction and explainable machine learning system.

Problem: Manual heart disease risk assessment is slow and doesn't scale — providers need a fast, reliable way to flag high-risk patients.

Solution: Built an end-to-end ML app predicting risk from 13 clinical features. Compared multiple models, tuned hyperparameters, and deployed a Random Forest model with Explainable AI (feature-importance) via an interactive Streamlit web app.

Result: A live, deployed tool evaluated on Accuracy, Precision, Recall, F1-Score & ROC-AUC — giving real-time, transparent risk predictions anyone can use.