SwiftSQL

SwiftSQL is a full-stack AI-powered application that translates natural language queries and ER diagrams into executable SQL queries. It combines deep learning models, a FastAPI inference service, a React frontend, and a Node.js backend with MySQL for seamless query generation, execution, and user session management.

---

Features

• Natural Language to SQL conversion via a sequence-to-sequence (LSTM) model.
• ER diagram parsing and conversion to database schema.
• Real-time SQL generation through FastAPI inference server.
• Interactive query and result interface built with React.
• Secure session-based login and SQL execution via Node.js and MySQL.
• Clean modular structure separating frontend, backend, and model layers.

---

Tech Stack

Layer	Technology
Frontend	React.js (Vite), HTML, CSS
Backend	Node.js, Express.js, MySQL
Model Server	FastAPI, TensorFlow, Pickle
Model Type	Sequence-to-Sequence (Encoder–Decoder) with LSTM
Auth System	Express Sessions
Tokenization	HuggingFace-style preprocessing with pickled tokenizers

---

Folder Structure

SwiftSQL/
├── fastapi-backend/       # FastAPI server for model inference
├── backend/               # Node.js backend with API and authentication
├── web-app/               # React-based frontend interface
├── LocalModel/            # Trained models and preprocessing tokenizers
│   ├── encoder_model.h5
│   ├── decoder_model.h5
│   ├── input_tokenizer.pkl
│   ├── target_tokenizer.pkl
│   ├── Max_lengths.pkl
│   └── sequence_params.pkl
├── app.py, utils.py       # Prediction logic for FastAPI server
├── server.js              # Express routes for login and SQL execution
├── Untitled.mdj           # (Optional) UML/ER diagram
├── package.json           # Project dependencies and metadata
└── .gitignore

---

Setup Instructions

1. Clone the Repository

git clone https://github.com/justbeingsniper/SwiftSQL.git
cd SwiftSQL

---

2. Set Up the FastAPI Server

cd fastapi-backend
pip install -r requirements.txt
uvicorn app:app --reload --port 8000

Ensure the following model files exist in fastapi-backend/models/:

• encoder_model.h5
• decoder_model.h5
• input_tokenizer.pkl
• target_tokenizer.pkl
• Max_lengths.pkl

---

3. Set Up Node.js Backend

cd ../backend
npm install
node server.js

Create a .env file in the backend/ directory:

DB_HOST=localhost
DB_USER=root
DB_PASSWORD=your_password

---

4. Start the React Frontend

cd ../web-app
npm install
npm run dev

Runs at http://localhost:5173.

---

API Endpoints

FastAPI (localhost:8000)

• POST /generate-sql — Converts natural language to SQL.

Node.js (localhost:5000)

• POST /api/login — User authentication.
• GET /api/databases — Lists MySQL databases.
• POST /api/execute — Executes a raw SQL query.

---

Example Usage

User query:
"Show all customers who placed more than 3 orders in the last month."

Workflow:

1. React app sends query to FastAPI.
2. FastAPI returns SQL:
   SELECT * FROM customers WHERE order_count > 3 AND order_date >= CURDATE() - INTERVAL 1 MONTH;
3. Node.js backend executes SQL and returns results to frontend.

---

License

This project is currently unlicensed.

---

Acknowledgements

• TensorFlow, Keras — Deep Learning Framework
• FastAPI + Uvicorn — Python-based model serving
• React + Vite — Frontend interface
• Express.js + MySQL — Backend and database layer
• Spider and WikiSQL — NLP-SQL training datasets

---

Author

Developed by justbeingsniper