DB-Unchained 🧠🔗

Turn plain English into database queries using AI with Auto Language Detection.

Example:

Prompt: Fetch all names and emails of users aged 25 or below from users table

Output:

SELECT name, email FROM users WHERE age <= 25;

---

🚀 Why DB-Unchained?

Tired of writing raw SQL or remembering Mongo filters? Just describe what you want — DB-Unchained translates your intent into a query.

---

🛠️ Features

• ✅ Converts English prompts to queries using LLM
• 🔍 Auto Language Detection Mode (Supports all existing Query Languages)
• 🌐 Support for multiple query languages in Manual Language Selection Mode (SQL or MongoDB)
• 🔁 Pluggable LLMs (OpenAI, Groq etc.)
• 🔧 Configurable via environment variables
• 📦 Dockerized for easy deployment
• ✨ Clean and modular structure, built for scaling

---

🧪 Example Prompts to Try

Prompt	Output Example
List all customers who joined in the last 30 days	SELECT * FROM customers WHERE created_at >= NOW() - INTERVAL 30 DAY;
Fetch products that are out of stock	SELECT * FROM products WHERE stock = 0;
Get users from New York aged between 20 and 30	SELECT * FROM users WHERE city = 'New York' AND age BETWEEN 20 AND 30;
Show total revenue for each product category	SELECT category, SUM(revenue) FROM products GROUP BY category;
Find all employees who haven't submitted their timesheet	SELECT * FROM employees WHERE id NOT IN (SELECT employee_id FROM timesheets);

---

⚙️ Configuration

All config is managed via environment variables. Example .env:

LLM_PROVIDER=groq
ENABLE_AUTO_DETECT=true
GROQ_API_KEY=your-groq-api-key-here
GROQ_MODEL=model-you-want-to-use

• ✅ Node.js v22+ supported natively with .env file alternatives
• 🔐 Never commit your .env with secrets

---

🔐 How to Get Groq API Key

• Visit: https://console.groq.com/keys
• Login with GitHub/Google
• Generate a new API key
• Add to .env as GROQ_API_KEY=...

---

▶️ Run Locally

# Install deps
npm install --omit=dev

# Run
npm start

• (if ENABLE_AUTO_DETECT = true) You’ll be prompted to enter your preferred query language followed by a plain text in your natural language. The app will return a raw DB query (acc to your selected query language in first prompt) using the selected LLM.

• (if ENABLE_AUTO_DETECT = false) You’ll be prompted to enter a plain text in your natural language. The app will return a raw DB query with automatic language selection based on your prompt and general popularity.

---

🐳 Docker

# Build
docker build -t db-unchained .

# Run with env
docker run -it --env-file .env db-unchained

• Make sure .env contains your API key and configs.

---

🤝 Contributing

PRs welcome. Just don’t break the modularity. Keep things clean, minimal, and extendable.

🧑‍💻 Author

Made with ❤️ by zlatanned.