This n8n workflow enables seamless voice and text communication via Telegram by integrating speech-to-text, AI responses, and text-to-speech conversion. It captures incoming messages (voice or text), transcribes audio using Deepgram, processes inquiries with OpenAI, retrieves relevant information from a Supabase knowledge base, and responds appropriately in voice or text form. Ideal for businesses aiming to provide intelligent, real-time customer support through natural language interactions.

Node Count	11 – 20 Nodes
Nodes Used	@n8n/n8n-nodes-langchain.agent, @n8n/n8n-nodes-langchain.lmChatOpenAi, @n8n/n8n-nodes-langchain.memoryBufferWindow, httpRequest, if, set, stickyNote, supabaseTool, telegram, telegramTrigger