WhatsApp Lead Bot (Twilio + Claude + Airtable)
Every inbound WhatsApp message is classified by Claude (pricing, demo, support), saved as a lead in Airtable and answered with the right reply – in seconds.
Download workflow (.json) Watch the video
What you need
n8n, Twilio WhatsApp, Anthropic Claude, Airtable. Pay per message (Twilio) + a few cents per 1,000 classifications (Claude Haiku). Self-hosted n8n or n8n Cloud.
Setup in 6 steps
- Import the JSON file in n8n: Workflows → Import from File.
- Airtable: create a base with a table "Leads" and the fields Phone, Name, Message, Intent, Received. Put your base ID (app…) and table ID (tbl…) into the "Save lead in Airtable" node and connect an Airtable token credential.
- Claude: add an Anthropic API credential to the "Claude: classify intent" node.
- Twilio: connect a Twilio credential. For testing, use the WhatsApp Sandbox number; in production, use your approved WhatsApp sender.
- Copy the production URL of the "WhatsApp message in" webhook and paste it into Twilio as the "When a message comes in" URL (HTTP POST). Activate the workflow.
- Replace the example reply texts and links (yourcompany pricing / booking pages) in the four Reply nodes.
Good to know
- The Twilio WhatsApp Sandbox only reaches people who first send your join code – fine for testing, not for real customers.
- WhatsApp only allows free-form replies within 24 hours after the customer's last message.
Free to use and adapt for your business. Provided as is, without warranty – see our terms. Prices and limits of third-party services change; check them before going live.