This is a lightweight, open-source demo that lets you ask natural language questions and get SQL queries in return — powered by Vanna.AI, OpenAI GPT, and a simple Flask UI.

Ask questions like:

• "Show all categories of sales"
• "List top-selling categories by revenue"
• "Get total orders per region"

✅ Natural language → SQL conversion
✅ Connects to a real SQLite database
✅ Powered by OpenAI GPT-4 + ChromaDB
✅ Flask-based interactive web interface
✅ Environment-based API key loading with .env
✅ Easily extensible to Postgres, MySQL, BigQuery, etc.

Vanna-Demo/
├── .env                # Your OpenAI API key (secure, ignored by Git)
├── db.py               # DB config for SQLite
├── demo_sales.db       # Sample SQLite database
├── main.py             # CLI + Flask app entry point
├── vanna_setup.py      # Vanna LLM setup and training
├── requirements.txt    # Python dependencies
└── readme.md           # This file

git clone https://github.com/your-username/Vanna-Demo.git
cd Vanna-Demo

pip install -r requirements.txt

Create a .env file in the root directory with your OpenAI API key:

OPENAI_API_KEY=sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXX

⚠️ This file is ignored by Git. Keep your API key private.

To run the app and open the Flask UI:

python main.py

Visit your browser at: http://127.0.0.1:5000

• "Show all customers who ordered more than once"
• "List total sales by region"
• "Find the average order value"

• Vanna.AI
• OpenAI GPT-4
• Flask
• ChromaDB
• SQLite

1. Loads table definitions from SQLite using introspection.
2. Trains the Vanna model with DDL, sample SQL, and documentation.
3. Accepts natural language questions via CLI or Flask UI.
4. Converts questions into SQL using OpenAI and your schema context.

• Add support for more databases (Postgres, MySQL, Snowflake, etc.)
• Enhance Flask UI with chat history or result charts
• Deploy to Docker or cloud

Created by [Pradeep Kumar]
Powered by Vanna.AI

MIT License