Built and deployed a production AI chatbot using Retrieval-Augmented Generation (RAG) to help enterprise users search and interact with internal knowledge. The system uses a fine-tuned Llama model combined with semantic search powered by Elasticsearch and MySQL. I designed the full pipeline including document ingestion, chunking strategy, vector similarity search, and LLM response generation. The chatbot integrates with existing SaaS systems and provides accurate contextual responses by retrieving relevant data before generating answers.