Telegram AI Assistant: Voice/Photo/Notes/Long-Term Memory

This n8n workflow creates a comprehensive Telegram personal assistant that processes voice messages, photos, and text inputs using AI. Incoming messages are routed dynamically: voice is transcribed, photos analyzed for details (with captions), and text processed directly. The GPT-4o-mini AI agent interprets content, recalls conversation history from PostgreSQL, and generates intelligent responses.

Key features include long-term memory and note-taking: the AI identifies noteworthy info to save as structured memories or concise notes in Baserow database tables. This enables persistent recall of personal details, reminders, and insights across sessions, making it ideal for personal productivity or finance tracking (e.g., expense notes from voice/photo).

Benefits: Saves hours on manual note-taking, enhances Telegram as a smart hub, scales for business use like client interactions. Use cases: Daily journaling via voice, photo-based expense logging, reminders, or custom finance assistants remembering user portfolios/preferences. Highly customizable with defined AI rules for memory/note decisions.