Shipped: Enterprise Knowledge Assistant — a RAG-based Q&A system

Built an Enterprise Knowledge Assistant — a RAG-based document search assistant that answers natural language questions using company policy documents, with source citations for every answer. Pipeline: documents are chunked → embedded with Gemini's embedding model → stored in ChromaDB → semantic retrieval on query → Gemini generates grounded, cited answers (with hallucination checks via prompt design). Backend: FastAPI, deployed on Render Frontend: HTML/CSS/JS, deployed on GitHub Pages Live demo: https://vedikaa17.github.io/enterprise-knowledge-assistant/ GitHub: https://github.com/Vedikaa17/enterprise-knowledge-assistant

0:00 / 0:00

Add a comment

0/2000