🚀 AI Chat-to-SQL Agent (n8n + PostgreSQL + OpenAI)
📌 Overview
This project demonstrates how to build a production-ready AI-powered data assistant that allows users to interact with a PostgreSQL database using natural language.
Instead of writing SQL queries manually, users can simply ask questions like:
"Show all products" "Which items are low in stock?" "What is the most expensive product?"
The system intelligently converts these queries into SQL, executes them, and returns structured, human-readable responses.
🧠 Key Concept
This project implements a Chat-to-SQL pipeline, where:
User input is captured via n8n chat trigger OpenAI model interprets the query AI generates a valid SQL statement PostgreSQL executes the query Results are formatted and returned to the user
🏗️ Architecture
⚙️ Tech Stack n8n – Workflow automation & orchestration PostgreSQL – Relational database pgAdmin – Database management UI OpenAI API – Natural language processing Docker – Containerized environment
📂 Project Structure . ├── docker-compose.yml ├── Chat with Database.json ├── README.md
🧩 Features Natural language to SQL conversion AI-powered query understanding Memory-enabled contextual conversations Secure query constraints (SELECT-only execution) Modular and extensible workflow design
🔐 Safety Controls The AI agent is constrained with rules to:
Only access the products table Execute SELECT queries only Prevent destructive operations (INSERT, UPDATE, DELETE, DROP)
🛠️ Setup Instructions
-
Clone the repository git clone https://github.com/your-username/chat-to-sql-ai-agent.git cd chat-to-sql-ai-agent
-
Start services using Docker docker-compose up -d
-
Access services n8n → http://localhost:5678 pgAdmin → http://localhost:5050
-
Configure PostgreSQL
Create table:
CREATE TABLE products ( id SERIAL PRIMARY KEY, name VARCHAR(255), price NUMERIC, stock INTEGER, category VARCHAR(50), created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP );
Insert sample data:
INSERT INTO products (name, price, stock, category) VALUES ('Laptop', 1200, 10, 'Electronics'), ('Mouse', 25, 200, 'Accessories'), ('Keyboard', 75, 150, 'Accessories'), ('Monitor', 300, 50, 'Electronics'), ('Office Chair', 180, 20, 'Office');
-
Import n8n Workflow Open n8n Import Chat with Database.json Configure: OpenAI API credentials PostgreSQL credentials
-
Activate Workflow
Start querying your database via chat.
💬 Example Queries “List all products” “Show low stock items” “Sort electronics by price” “How many products per category?”
📈 Use Cases Business intelligence assistant HR analytics chatbot Inventory management insights Data exploration without SQL knowledge
🚀 Future Enhancements Multi-table support Role-based query access Dashboard integration (Streamlit / React) API exposure for external apps Vector search + semantic querying
🤝 Contribution
Contributions, improvements, and suggestions are welcome.
📄 License
MIT License