Query PostgreSQL database with natural language using Groq AI chatbot

This guide shows you how to deploy a chatbot that lets you query your PostgreSQL database using natural language. You will build a system that accepts chat messages, retains conversation history, constructs dynamic SQL queries, and returns responses generated by an AI model. By following these instructions, you will have a working solution that integrates n8n’s AI Agent capabilities with PostgreSQL.


Prerequisites

Before you begin, ensure that you have the following:

An active n8n instance (self-hosted or cloud) running version 1.50.0 or later. Valid PostgreSQL credentials configured in n8n. API credentials for the Groq Chat Model (or your preferred AI language model). Basic familiarity with SQL (specifically PostgreSQL syntax) and n8n node concepts such as chat triggers and memory buffers. Access to the n8n Docs on AI Agents for further reference.

Workflow Setup

1. Chat Interface & Trigger When Chat Message Received: This node listens for incoming chat messages via a webhook. When a message arrives, it triggers the workflow immediately.

2. Conversation Memory Chat History: This memory buffer node stores the last 10 interactions. It supplies conversation context to the AI Agent, ensuring that responses consider previous messages.