A sophisticated web application leveraging Retrieval-Augmented Generation (RAG) to enhance user exploration of technical documentation knowledge bases.

Key Achievements:

• Advanced LLM Integration: Implemented cutting-edge language models for query refinement (subquery generation), cross-encoder scoring for document relevance ranking, and ultimately, contextually informed answer generation.

• Chroma Vector Database: Utilized ChromaDB as a high-performance vector database to store and retrieve document embeddings, enabling efficient semantic search within the knowledge base.

• PDF Ingestion Pipeline: Designed a robust process to ingest and preprocess technical documentation in PDF format, transforming unstructured data into valuable insights.

• Full-Stack Development Expertise: Developed the entire application stack, using React for the frontend, Flask for the backend, PostgreSQL for logging, and Docker for seamless containerization.

• Customizable Logging: Built a Python-based logger to capture errors and user interactions, providing valuable data for analysis and improvement.