n8n: WhatsApp Chat Agent with RAG скачать в хорошем качестве

n8n: WhatsApp Chat Agent with RAG

I built a WhatsApp chat agent on n8n that functions as a sales assistant or customer support representative using a RAG workflow. Internal documents are stored in a Supabase vector database after being split and converted into embeddings. Incoming messages, text or audio, are processed, transcribed if needed, and sent to GPT-4o Mini, which generates replies using relevant knowledge from the vector store and chat history. The workflow ensures accurate, context-aware responses while maintaining conversation memory and fully automates chat handling.