AI Document Q&A Agent

RAG-based document Q&A app — upload PDFs/text, ask questions, get answers with chunk-level citations. Runs in mock mode without paid LLM keys.

Python 3.11FastAPIReact + ViteRAGLexical retrievalCitationsDocker Compose

View repository

Pattern

Retrieve · Cite · Generate

Run mode

Mock or OpenAI-backed

Project Overview

FastAPI + React product: upload panel, document library, question workbench, answer panel, and citation list. Pipeline covers ingestion, chunking, lexical retrieval, citation builder, and a swappable mock/OpenAI answer generator.

Challenge

Generic chatbots hallucinate and provide no audit trail back to source material — unusable for any document-grounded use case.

Solution

Built an RAG pipeline that separates retrieval from generation, with chunk-level citations attached to every answer and a deterministic mock generator so reviewers can run the demo without API keys.

Results

Reproducible portfolio MVP that demonstrates real RAG product thinking — citations, evaluation surface, and a swap-in path to production LLMs.