Fetches web content and returns relevant chunks using RAG with semantic chunking, query-based scoring, and LexRank, supporting PDFs and 50+ languages.
