🗄️ Natural Language to SQL Generator

An AI-powered app that converts plain English questions into SQL queries and executes them against a SQLite database.

🚀 Features

• Natural language to SQL using Groq (Llama 3.1)
• Auto-executes queries against SQLite
• Results shown as sortable table with CSV download
• Plain English explanation of results
• 10 example questions to try instantly
• Query history panel

🛠️ Tech Stack

• LLM: Llama 3.1 8B via Groq API (free)
• Database: SQLite with SQLAlchemy
• Backend: LangChain + Python
• Frontend: Streamlit

⚙️ Setup

1. Clone the repo

git clone https://github.com/YOUR_USERNAME/text-to-sql.git
cd text-to-sql

2. Install dependencies

pip install -r requirements.txt

3. Get a free Groq API key

• Go to console.groq.com
• Sign up with Google
• Create an API key

4. Create .env file

GROQ_API_KEY=gsk_xxxxxxxxxxxxxxxxxxxxxxxx

5. Create the sample database

python sample_db.py

6. Run the app

python -m streamlit run app.py

Open http://localhost:8501

💡 Example Questions

• "Show me the top 5 most expensive products"
• "Which customer has spent the most money?"
• "What is the total revenue per product category?"
• "Show me all orders placed in March 2024"

📁 Project Structure

text-to-sql/
├── app.py            # Streamlit frontend
├── sql_pipeline.py   # SQL generation + execution pipeline
├── sample_db.py      # Creates demo SQLite database
├── requirements.txt  # Dependencies
├── .env              # API keys (not committed)
└── README.md         # This file