Graph RAG Flask Application

A Retrieval-Augmented Generation (RAG) application using Knowledge Graphs (KùzuDB), Vector Search (FAISS), and Local LLMs (Ollama/DeepSeek).

Features

• Hybrid Retrieval: Combines vector similarity search with knowledge graph traversal.
• Local Processing: Uses local LLMs via Ollama for privacy and cost efficiency.
• Graph Visualization: Interactive graph view using Cytoscape.js.
• Document Support: PDF, DOCX, TXT, MD.

Prerequisites

1. Install Ollama

Download and install Ollama from ollama.com.

2. Pull the Model

This project uses deepseek-r1:8b by default. Run the following command in your terminal:

ollama pull deepseek-r1:8b

Start the Ollama server:

ollama serve

(Note: On some systems, the desktop app runs the server automatically)

3. Python Environment

Ensure you have Python 3.9+ installed.

Installation

1. Clone or navigate to the project directory:

   cd graphrag-flask

2. Create a virtual environment (optional but recommended):

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # Mac/Linux
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

Configuration

The application uses a config.py file which loads environment variables. You can create a .env file to override defaults:

OLLAMA_HOST=http://localhost:11434
OLLAMA_MODEL=deepseek-r1:8b
KUZU_DB_PATH=./kuzu_db
CHUNK_SIZE=500

Running the Application

1. Start the Flask server:

   python app.py

2. Open your browser and navigate to: http://localhost:5000

Usage

1. Upload: Use the sidebar to upload a document.
2. Process: Click "Process" to extract entities and build the graph.
3. Chat: Ask questions about your documents in the chat window.