Multi-Modal Telegram AI Assistant with Gemini, Voice & Image Gen

This n8n workflow creates an advanced multi-modal AI assistant named Simran for Telegram, capable of processing text and voice messages. It transcribes voice via AssemblyAI, analyzes intent with a Large Language Model (likely Gemini), fetches conversation context from MongoDB, and routes to specialized tasks: general chat with customizable personality (standard or 'Girlfriend Mode') and mood analysis, image generation from prompts, reminder setting via Google Calendar and Tasks, or saving user info to memory.

Key benefits include seamless multi-modal interaction, persistent context for natural conversations, automation of reminders and content generation, saving hours of manual setup for AI bots. It's highly customizable, toggling AI personalities and integrating premium services for professional-grade assistance.

Ideal use cases: Personal AI companions for daily tasks, customer support bots with voice/image capabilities, productivity tools for reminders and memory, or experimental agentic AI prototypes. Perfect for developers building Telegram ecosystems, reducing development time from weeks to hours.