🚀 SQL Generator AI (Spring AI + Groq)

Generate SQL queries from natural language using LLMs
Built with Spring Boot • Powered by Groq • Production-ready architecture

---

📌 Overview

SQL Generator AI is a backend service that transforms natural language queries into SQL statements using Large Language Models (LLMs) via Spring AI.

This project reflects real-world backend engineering practices including:

• Integration of modern LLM infrastructure in Spring Boot
• Clean and scalable layered architecture
• Effective prompt engineering with schema grounding
• Production-oriented API design

---

🧠 How It Works

Client Request → REST Controller → Service Layer → Spring AI → Groq (LLaMA 3) → SQL Response

1. The client sends a natural language query
2. The backend enriches the prompt with a predefined schema
3. The request is processed via Spring AI's ChatClient
4. Groq LLM generates an optimized SQL query
5. The API returns the SQL response

---

🤖 AI Generated SQL Result

---

---

✨ Key Features

• Natural Language → SQL query generation
• Integration with Groq (LLaMA 3.3 70B model)
• Built using Spring AI ChatClient
• Clean controller-service architecture
• Schema-aware prompt design for higher accuracy
• Lightweight, extensible, and production-ready

---

🏗️ Project Structure

sql-generator-ai/
│
├── src/main/java/com/example/sqlgeneratorai/
│   ├── controller/
│   │   └── SqlController.java
│   ├── service/
│   │   └── SqlGeneratorService.java
│   ├── dto/
│   │   └── SqlRequest.java
│   └── SqlGeneratorAiApplication.java
│
├── src/main/resources/
│   └── application.properties
│
├── src/test/
├── pom.xml
└── mvnw

---

⚙️ Tech Stack

• Language: Java 21
• Framework: Spring Boot
• AI Integration: Spring AI
• LLM Provider: Groq
• Model Used: LLaMA 3.3 (70B Versatile)
• API Style: RESTful APIs
• Build Tool: Maven
• Development Environment: VS Code

---

📡 API Endpoint

🔹 Generate SQL

POST /api/sql/generate

📥 Request

{
  "question": "Get all employees with salary greater than 50000"
}

📤 Response

SELECT * FROM employees WHERE salary > 50000;

---

🧠 Prompt Engineering

The application uses a predefined schema to guide the model:

TABLE employees(
    id BIGINT,
    name VARCHAR(100),
    salary DECIMAL,
    department_id BIGINT
);

Benefits

• Improves SQL accuracy
• Reduces hallucinated queries
• Ensures schema consistency

---

⚙️ Configuration

Configure your application in application.properties:

spring.application.name=sql-generator-ai

spring.ai.openai.api-key=YOUR_API_KEY
spring.ai.openai.base-url=https://api.groq.com/openai
spring.ai.openai.chat.options.model=llama-3.3-70b-versatile

---

🚀 Getting Started

📥 Clone Repository

git clone https://github.com/SRCarlo/sql-generator-ai.git
cd sql-generator-ai

---

🏗️ Build Project

mvn clean install

---

▶️ Run Application

mvn spring-boot:run

---

🌐 Test API

POST http://localhost:8080/api/sql/generate

---

🧪 Testing

mvn test

---

🔍 Code Highlights

Controller Layer

• Exposes REST endpoint /api/sql/generate
• Handles incoming HTTP requests

Service Layer

• Uses Spring AI ChatClient
• Injects schema into prompt
• Processes LLM response

DTO Layer

public record SqlRequest(String question) {}

---

📈 Future Enhancements

• Dynamic schema input support
• SQL validation and optimization layer
• Multi-database support (MySQL, PostgreSQL, etc.)
• Swagger / OpenAPI documentation
• Authentication & rate limiting
• Docker containerization

---

⚠️ Security Best Practices

• Never expose API keys in source code
• Use environment variables or secret managers
• Implement request validation and rate limiting

---

🤝 Contributing

Contributions are welcome!

git checkout -b feature/your-feature
git commit -m "feat: add your feature"
git push origin feature/your-feature

---

🛡️ License

This project is licensed under the MIT License.

You are free to use, modify, and distribute this software with proper attribution.

---

👨‍💻 Author

Shubham Raut
Backend Developer | Java | Spring Boot | AI Integrations

---

⭐ Support

If you found this project helpful:

• ⭐ Star the repository
• 🍴 Fork the project
• 📢 Share it with others

---

Built with ❤️ using Spring AI and LLMs