NFL AI Draft Scout

Project Description

The NFL Draft Scout AI is a full-stack RAG (Retrieval-Augmented Generation) application that combines semantic search with AI-powered scouting analysis for 500 prospects in the 2026 NFL draft class. The application features a conversational chatbot interface where users can query prospects using natural language, with the system retrieving relevant candidates from a ChromaDB vector database and generating contextual analysis through Claude AI that synthesizes player statistics, consensus rankings, and performance metrics. Built with a React/Tailwind frontend and Python Flask backend deployed on Railway, the system implements a hybrid query strategy that combines sentence-transformer embeddings (all-MiniLM-L6-v2) for semantic similarity with statistical re-ranking algorithms to surface the most relevant prospects. The project showcases advanced vector database architecture with performance contextualized prospect descriptions optimized for semantic search accuracy, temporal ranking tracking infrastructure to monitor draft stock momentum, and intelligent data integration handling 377 prospects with complete statistical coverage while gracefully managing positions where traditional statistics don't apply (offensive linemen, specialists), demonstrating practical implementation of modern AI retrieval systems for domain-specific analytical applications.