WhatsApp AI Customer Support Agent (n8n + Claude + WhatsApp API)

We run an e-commerce store and our support team is drowning in repetitive WhatsApp messages — order status, return policy, product questions, opening hours. The same 10 questions, all day, every day. We need an AI agent that handles these automatically so our team can focus on the cases that actually need a human. What we need: An AI support agent connected to our WhatsApp Business number that reads incoming messages, understands what the customer is asking, and replies instantly with the right answer. It should pull from our product/policy knowledge base so answers are accurate, not generic. If it can't handle something, it should flag it and hand off to a human agent cleanly. Specifically: Respond to common queries instantly: order status, returns, shipping times, product info, opening hours Pull answers from our knowledge base (FAQs, policy docs, product catalogue) Detect when a query is too complex or sensitive and route it to a human with context Log every conversation so we have a full record Easy to update — we should be able to add new FAQs without a developer Deliverables: Working n8n workflow connected to WhatsApp Business API (via Twilio or Meta Cloud API) Claude-powered response layer using our knowledge base (RAG) Human handoff logic with escalation trigger Conversation logging (Google Sheets or Airtable) Short walkthrough so we can maintain the FAQ ourselves Skills: n8n, Claude API, WhatsApp Business API, RAG, Retrieval Augmented Generation, Conversational AI, Twilio, Meta Cloud API, Google Sheets, Workflow Automation, Chatbot, Customer Support Automation