title	OmniDoc QA
emoji	🧠
colorFrom	blue
colorTo	green
sdk	docker
app_port	7860

OmniDoc QA — Multi-Modal RAG System

A production-style RAG (Retrieval-Augmented Generation) application that lets users upload PDFs or websites, then ask natural-language questions over the ingested content. Built with Flask, LangChain, ChromaDB, and Groq.

🚀 Try it live on Hugging Face Spaces: OmniDoc QA

Features

• Multi-format ingestion — Upload PDFs or paste a website URL; text, tables, and structure are extracted automatically
• RAG pipeline — Documents are chunked, embedded with all-MiniLM-L6-v2, stored in ChromaDB, and retrieved at query time
• Comparative queries — Detects cross-document questions and balances retrieval across sources
• Inline metrics — Each response shows latency, chunks retrieved, and token usage
• Feedback loop — Thumbs up/down per response, stored in a Hugging Face Dataset for analysis
• Monitoring dashboard — Streamlit dashboard pulls logs from HF Dataset to visualize latency trends, token usage, and retrieval quality

Quickstart

# Clone and install
git clone https://github.com/aniketpoojari/OmniDoc-QA-LLM-RAG.git
cd OmniDoc-QA-LLM-RAG
pip install -r requirements.txt

# Set API keys
echo "GROQ_API_KEY=your_key_here" >> .env
echo "HF_TOKEN=your_hf_write_token" >> .env

# Run
python app.py
# Open http://localhost:7860

Monitoring Dashboard

streamlit run dashboard.py

Pulls logs from the HF Dataset and displays latency trends, token costs, retrieval quality, and query history.

Tech Stack

Layer	Technology
Backend	Flask
LLM	Groq API (LLaMA 3.1 8B Instant)
Embeddings	HuggingFace all-MiniLM-L6-v2
Vector DB	ChromaDB
Orchestration	LangChain
PDF extraction	PyMuPDF + Tabula
Web extraction	BeautifulSoup4
Frontend	Bootstrap 5, jQuery, DOMPurify
Logging	Hugging Face Datasets (via HfApi)
Deployment	Docker on Hugging Face Spaces
CI/CD	GitHub Actions

Project Structure

├── app.py                          # Flask routes and main entry point
├── models/
│   └── vector_store.py             # ChromaDB wrapper (chunk, embed, store)
├── services/
│   ├── llm_service.py              # Groq LLM integration and RAG chain
│   ├── pdf_extraction_service.py   # PDF text + table extraction
│   ├── website_extraction_service.py # Web scraping and content extraction
│   └── monitoring_service.py       # HF Dataset logging and feedback
├── templates/
│   └── index.html                  # Main UI template
├── static/
│   ├── script.js                   # Frontend logic (AJAX, chat, feedback)
│   └── style.css                   # Custom styles
├── dashboard.py                    # Streamlit monitoring dashboard
├── Dockerfile                      # HF Spaces deployment
├── .github/workflows/
│   └── deploy-hf-spaces.yml        # CI/CD: GitHub → HF Spaces
└── requirements.txt

See SYSTEM_DESIGN.md for architecture details and design decisions.

Environment Variables

Variable	Required	Description
GROQ_API_KEY	Yes	Groq API key for LLM inference
HF_TOKEN	Yes	HuggingFace write token for logging and deployment
PORT	No	Server port (default: 7860)