Build n8n AI WhatsApp Agent

I have a clear goal: launch an AI-driven WhatsApp agent built entirely in n8n that can chat naturally with customers, qualify leads, answer product questions, recommend items, follow up, and trigger sales automations without human intervention. Here’s the environment I envision • n8n will act as the orchestration layer. Your flows should pull in an NLP engine of your choice through n8n’s HTTP or pre-built OpenAI / Dialogflow nodes so intent detection and response generation happen seamlessly. • The agent must cover three core use-cases from day one—customer handling, lead qualification, and product queries—while remaining easy to expand into follow-ups and revenue tracking later. • I’m not yet certain whether our WhatsApp Business API is active. I’ll need you to audit our current setup, advise on any missing Meta Business Manager steps, and connect the final number to n8n once we’re ready. Success looks like 1. A fully documented n8n workflow (export JSON) that includes trigger, NLP processing, branching logic, and CRM/webhook actions. 2. Tested live on a sandbox or approved number, showing accurate intent recognition and correct data capture for the three priority tasks. 3. Clear hand-off notes so my internal team can tweak prompts, add steps, or migrate to production with minimal guidance. If you’ve already built conversational WhatsApp automations in n8n, you’ll know the best nodes, error-handling tricks, and how to keep response times lightning-fast. That’s exactly the expertise I need to get this into customers’ hands quickly.