n8n WhatsApp FAQ Webhook: OpenAI Fallback by Keyword Matching

Instant WhatsApp FAQ replies with keyword matching—and a safe OpenAI fallback

This n8n workflow listens to WhatsApp Business Cloud API webhook events, matches incoming text messages to your FAQ list using keyword matching, and—when no direct match is found—uses an OpenAI fallback that can only return results from approved FAQ IDs before replying back to the customer.

What this workflow does

• Verifies the webhook callback URL: Handles the Meta verification GET request and returns hub.challenge only when the provided verify token matches.
• Processes incoming WhatsApp messages: Accepts webhook POST events, keeps only incoming text messages, and ignores delivery/read receipts.
• Answers via keyword-matched FAQs: Scores each message against a built-in FAQ list and selects the highest-scoring FAQ answer when a match is found.
• Uses OpenAI when no keyword match exists: Sends the customer message plus the available FAQ IDs/questions to an OpenAI chat model to select a single FAQ ID or null.
• Ensures safe output: Validates the model’s result against known FAQ IDs and replies with the stored answer—or a fixed fallback message—if no valid selection is made.
• Sends the reply back on WhatsApp: Uses the same phone number that received the original message.

Use cases

• Customer support for SaaS operators needing fast WhatsApp responses to common questions.
• FAQ-driven lead qualification where users ask about pricing, availability, or features.
• Teams that want deterministic keyword FAQ behavior, with OpenAI as a controlled “no-match” resolver.

Technical details

• WhatsApp Business Cloud API: Webhook verification and message reply handling.
• n8n nodes: webhook, if, code, respond to webhook, and workflow notes via sticky note.
• OpenAI / OpenAI-compatible model: Used only for FAQ ID selection with validation against the approved FAQ list.