Document Intelligence Platform

A RAG (Retrieval-Augmented Generation) system for intelligent document analysis. Designed to handle complex documents, tables, and reasoning.

Key Features

• RAG Architecture: Retrieves relevant chunks from financial documents to answer user queries.
• Hybrid Search: Combines semantic search (MiniLM) with keyword matching (BM25) for precision.
• Citations: Answers include specific page numbers and source text snippets.

Quick Start

Installation

# Clone the repository
git clone https://github.com/Nashid-Noor/financial-doc-intelligence.git
cd financial-doc-intelligence

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Linux/Mac
# or
.\venv\Scripts\activate  # Windows

# Install dependencies
pip install -r requirements.txt

# Setup Environment
cp .env.example .env
# Edit .env and add your HF_API_KEY

Running the Application

Option 1: Quick Start (Recommended)

The easiest way to run everything (API + UI) is using the helper script:

# Make executable first
chmod +x run.sh

# Start everything
./run.sh all

Option 2: Manual Start

cd src/api
uvicorn app:app --host 0.0.0.0 --port 8000 --reload

2. Start the Streamlit UI

streamlit run ui/streamlit_app.py

3. Access the Application

• Streamlit UI: http://localhost:8501
• API Docs: http://localhost:8000/docs
• API ReDoc: http://localhost:8000/redoc

Configuration

RAG Configuration (configs/rag_config.yaml)

retrieval:
  embedding_model: "sentence-transformers/all-MiniLM-L6-v2"
  chunk_size: 512
  chunk_overlap: 50

Development

Running Tests

# All tests
pytest tests/ -v

# With coverage
pytest tests/ --cov=src --cov-report=html

Docker Deployment

# Build image
docker build -t financial-doc-intelligence .

# Run container
docker run -p 8000:8000 -p 8501:8501 financial-doc-intelligence

Render Deployment

The project includes a render.yaml file for easy deployment on Render.com.

1. Create a Key-Value Secret file or simple environment variable for HF_API_KEY.
2. Connect your repository to Render.
3. Select "Web Service" and use the Docker runtime.