n8n WhatsApp Webhook AI Reply: Wafly + OpenAI Agent

Turn WhatsApp questions into instant, AI-powered replies (with per-contact memory)

This n8n workflow listens for incoming WhatsApp messages from Wafly, groups rapid-fire questions into one clean prompt (optionally transcribes voice notes), then uses an OpenAI-powered agent with per-phone-number memory to generate a short reply and send it back to the customer.

What this workflow does

• Receives WhatsApp webhooks through an n8n webhook endpoint exposed by the workflow.
• Normalizes the payload so either typed text or a voice-note transcription becomes a single question field.
• Filters non-customer events, including echoes from your own number, group messages, and empty messages—so the AI only responds to real customer input.
• Buffers rapid-fire messages to combine multiple messages into one question before calling the AI agent.
• Calls an OpenAI chat model via an AI agent and uses a memory window keyed to the customer’s phone number to maintain context.
• Sends the generated response back to the customer as a WhatsApp text message using Wafly.

Use cases

• SaaS support: answer repeated questions like pricing, setup steps, or troubleshooting without manual typing.
• Lead qualification: respond to incoming WhatsApp inquiries while maintaining context per contact.
• Sales & onboarding: handle multi-message questions as one consolidated request.
• Voice note handling: turn voice notes into text (optional) and reply automatically.

Technical details

• Self-host required: install the community node n8n-nodes-wafly (this template does not work on n8n Cloud).
• Integrations: Wafly webhook + Wafly sending; OpenAI API for the chat agent response.
• Workflow logic nodes: webhook, if, code, no op, plus n8n-nodes-wafly and supporting nodes (e.g., sticky note for guidance).
• Setup steps: configure Wafly credentials, add OpenAI credentials, run the manual buffering sequence, and register the workflow webhook URL with Wafly.