📡 NL2SQL API

A FastAPI-based backend that converts natural language questions into SQL queries, executes them on a SQLite database, and optionally returns results as downloadable Excel files.

---

🚀 Features

• 🧠 Natural Language to SQL conversion
• 🔍 SQL query execution on SQLite
• 📊 Return query results as JSON
• 📁 Excel download endpoint (.xlsx)
• 🌐 Designed to integrate with iOS (SwiftUI) chat apps

---

📦 API Endpoints

POST /query

Request:

{
  "question": "Show all employees"
}

** Response:**

{
  "query": "SELECT * FROM employees;",
  "results": [
    [1, "Alice", "HR", 60000],
    [2, "Bob", "IT", 75000]
  ]
}

Response:

Returns: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

Attachment: Excel file containing SQL results

📁 Project Structure

nl2sql-api/
├── main.py             # FastAPI entrypoint
├── model.py            # NL-to-SQL logic
├── schema.py           # Pydantic models
├── database.py         # DB setup and sample data
├── excel.py            # Excel export logic
├── requirements.txt    # Python dependencies
└── .renderignore       # Ignore unnecessary files during deployment

🧪 Local Setup

1. Clone the repo:

git clone https://github.com/YOUR_USERNAME/nl2sql-api.git
cd nl2sql-api

3. Create and activate a virtual environment:

python3 -m venv .venv
source .venv/bin/activate

4. Install dependencies:

pip install -r requirements.txt

6. Run locally:

uvicorn main:app --reload

🌍 Deploy

This app can be deployed to Render with:

Auto deploy on main branch push

render.yaml (optional)

.renderignore to skip local-only files

📱 NL2SQL Chat App

NL2SQL Chat App

📄 License

MIT License © 2025