Turn Telegram chatbots into GPT-4o-mini advisors with Postgres memory and Shopify data

How it works

• Receives a user message from a chat
• Uses an LLM node to understand user intent and generate a response
• Retrieves and stores conversation memory in a database (Supabase)
• Fetches real-time product or business data from an external API (e.g. Shopify)
• Applies custom logic to ensure a consistent, human-like tone
• Sends a contextual response back to the user

Set up steps

• Import the workflow into n8n
• Create and connect the required credentials (LLM, database, API)
• Configure environment variables and placeholders
• Adjust the prompt and logic to your specific business needs
• Test the workflow using sample user messages