Back to projects
Full Stack Developer

PDF Chat RAG

Retrieval-Augmented Generation workspace for PDF documents with multi-PDF upload, explicit status lifecycle, streaming chat responses, and source-backed answers.

Live DemoSource Code
Duration

2 months

Team

Solo

Status

Completed

Stack

3 Techs

Key Features
Multi-PDF upload and indexing
Streamed chat responses
Source citations with answers
Top-K retrieval control
Document-level filtering
pdf-chat-rag.app
PDF Chat RAG screen 1

The Challenge

  • •Existing PDF chat apps mix ingestion and chat in one path
  • •No clear document processing states
  • •Retrieval evidence is hidden from users

The Solution

  • Separated ingestion from chat pathways
  • Implemented status lifecycle: queued, processing, ready, failed
  • Added source snippets and similarity evidence

Impact & Learnings

  • •Learned RAG architecture fundamentals
  • •Built reliable async document processing
  • •Created inspectable retrieval system

Future Roadmap

  • •Add user authentication
  • •Implement chat history
  • •Multi-user support