A full-stack AI-powered music intelligence platform built to practice machine learning and Python. Features an XGBoost genre classifier trained on 89,000+ tracks across 125 genres, mood-based discovery, a vibe builder with custom audio feature controls, a BPM-phased workout playlist generator, and a hybrid recommendation engine combining collaborative filtering with Pinecone vector similarity search. Built with a FastAPI microservices backend, React 19 frontend, PostgreSQL database, and fully containerized with Docker Compose.