LLM Knowledge Retrieval System (RAG)

AI-powered document Q&A system that lets you upload PDFs and ask questions with source citations, using Retrieval-Augmented Generation (RAG).

---

Demo

Upload a PDF → Ask a question → Get an answer + cited snippets

(Add screenshots / short GIF here)

---

Key Features

• ✅ Upload PDF documents
• ✅ Ask natural-language questions
• ✅ Answers include source citations (document + chunk)
• ✅ Persistent vector store (data survives restart)
• ✅ Simple REST API (FastAPI)
• ✅ Optional lightweight web UI (if you built one)

---

Tech Stack

• Backend: FastAPI (Python)
• Vector DB: ChromaDB
• Embeddings: OpenAI text-embedding-3-small
• LLM: gpt-4o-mini

---

How It Works (RAG Pipeline)

Indexing

1. PDF → Text extraction
2. Chunking (split into small overlapping text blocks)
3. Embeddings (convert chunks into vectors)
4. Store vectors + metadata in ChromaDB

Querying

1. User question → Embedding
2. Similarity search in ChromaDB (top-k chunks)
3. Retrieved chunks → injected as context into the LLM prompt
4. LLM generates answer + citations

---

Future Improvements

• Support more file types (DOCX, TXT)
• Better chunking strategies
• Conversation memory
• Multiple document queries

---