The system is triggered by incoming messages from three distinct sources: WhatsApp, Telegram, and a general chat interface. All messages are routed to a central AI Agent, which is powered by Google Gemini Chat models, enhanced with Pinecone Vector Store for semantic search, and supported by Hugging Face embeddings for vector representation.

A Simple Memory module is used to maintain context and continuity across sessions, ensuring personalized interactions. The AI Agent intelligently decides when to query the vector store to retrieve relevant answers, then generates a contextual response.

A Switch node is used to route the AI's response back to the appropriate platform (WhatsApp or Telegram), ensuring seamless channel-specific delivery through WhatsApp Business Cloud or Telegram Bot API.

This solution demonstrates the orchestration of advanced AI components to create a unified smart agent capable of natural, context-aware communication across multiple platforms—suitable for enterprise customer support, chatbots, and intelligent digital assistants.