Project Overview
Objective
Built semantic document QA over websites and PDFs with hybrid LLM support.
Stack
PythonLangChainFAISSOllama (LLaMA 3.2)GPT-4o
Delivery highlights
- Developed a document questionanswering system using Python and LangChain to ingest and process content from websites and PDF files, with semantic retrieval powered by embeddings and FAISS. Integrated both local LLMs via Ollama (LLaMA 3.2) and cloud-based models (GPT-4o) to generate accurate, context-aware responses based on retrieved document content.