title	emoji	colorFrom	colorTo	sdk	pinned
Hydro Rag	🐢	indigo	green	docker	false

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Hydrogen Certification RAG System

A smart question-answering system for hydrogen certification documents using RAG (Retrieval-Augmented Generation).

What it does

• Ask questions about hydrogen certifications in plain English
• Get accurate answers from official certification documents
• Compare different chunking strategies (basic vs hybrid)
• Automatic certification detection from your questions

Quick Start

1. Setup Environment

# Clone the repository
git clone https://github.com/Goodnight77/hydro_rag
cd hydro_rag

# Create environment file
cp .env.example .env

2. Add API Keys

Edit .env file and add:

OPENAI_API_KEY=your_openai_key_here
GROQ_API_KEY=your_groq_key_here
ELASTICSEARCH_HOSTS=http://localhost:9200

3. Run with Docker

# Build and run
docker build -t hydrogen-rag .
docker run -p 7860:7860 hydrogen-rag

4. Use the App

Open http://localhost:7860 in your browser and start asking questions!

Example Questions

• "What are the purity requirements in GH2 Standard?"
• "How does CertifHy certification work?"
• "What are the safety protocols for hydrogen storage?"

Available Certifications

• GH2 Standard
• CertifHy (NGC & RFNBO)
• ISO 19880 Hydrogen Quality
• ISCC EU/PLUS/CORSIA
• REDcert-EU
• TUV Rheinland H2.21
• And more...

How it Works

1. Document Processing: Converts PDFs, DOCX, and XLSX files into searchable chunks
2. Smart Chunking: Uses both basic and semantic chunking strategies
3. Question Classification: Automatically detects which certification you're asking about
4. Hybrid Search: Combines text matching and vector similarity for better results
5. AI-Powered Answers: Uses LLMs to generate clear, accurate responses

Tech Stack

• Frontend: Streamlit
• Backend: FastAPI
• Search: Elasticsearch
• LLMs: Groq (Llama 3.3, Gemma2)
• Embeddings: OpenAI text-embedding-3-small
• Documents: PDF, DOCX, XLSX support

Development

Local Setup

# Install dependencies
pip install -r requirements.txt

# Start Elasticsearch
# (See Docker setup or install locally)

# Run the app
streamlit run streamlit.py

API Usage

# Start FastAPI server
uvicorn app:app --host 0.0.0.0 --port 8000

# Query the API
curl -X POST "http://localhost:8000/query" \
  -H "Content-Type: application/json" \
  -d '{"query": "What are GH2 purity requirements?"}'

Project Structure

├── app.py              # FastAPI backend
├── streamlit.py        # Streamlit frontend
├── chunking/           # Text chunking strategies
├── elastic/            # Elasticsearch setup
├── embeddings/         # OpenAI embeddings
├── prompting/          # LLM query processing
├── file_processing.py  # Document parsing
└── Dockerfile          # Container setup

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

---

Need help? Open an issue or check the documentation in each module.