• Developed a document-focused chatbot using Llama 3.2, BGE Embeddings, Qdrant allowing users to interact with their 100% data from 5 file types (PDF, TXT, DOCX, CSV, and PPTX). • Implemented and fine-tuned RAG (Retrieval-Augmented-Generation) using LangChain for efficient embedding generation and intelligent response generation, document retrieval from vector database via Docker. • Built an interactive streamlit-based UI, integrating file inputs and previews, generate embeddings, real-time chat.
