Telegram AI Bot Workflow: Supabase Memory + OpenAI

Build a Telegram AI bot with real memory (Supabase + OpenAI) — ready to run in n8n

This n8n workflow turns your Telegram bot into a context-aware assistant by combining OpenAI for responses and Supabase to store and retrieve user memory. The result: conversations stay continuous and personalized instead of starting over every message.

What this workflow does

• Receive New Message: The bot listens for incoming messages from users in Telegram.
• Check user in the database: It verifies whether the user exists by using identifiers like telegram_id and links them to an openai_thread_id.
• Answer with OpenAI Assistant: The workflow routes the user’s prompt to OpenAI to generate a response while preserving conversation context.
• Persist memory in Supabase: User/session data is stored in Supabase so the bot can reference past interactions and maintain continuity over time.

Use cases

• Customer support & FAQs: Keep context across messages so users don’t repeat themselves.
• Virtual assistants: Provide a more human-like experience by remembering prior requests and preferences.
• Chatbot enthusiasts & developers: Use a production-oriented pattern for Telegram bots with persistent memory.
• SaaS operators: Add continuity to your Telegram-based assistant without building a custom memory layer from scratch.

Technical details

• Integrations: Telegram, OpenAI, Supabase
• n8n nodes / components mentioned: if, merge, supabase, telegram, sticky note, http request
• Memory mechanism: Stores and retrieves user/session continuity using telegram_id and openai_thread_id

Included with the template is a detailed video guide showing how to build the bot from a simpler version to a more complex setup—so you can follow the full process end-to-end.