Deploy a Live AI Assistant on WhatsApp to Answer Customer Questions

Deploy a smart AI chatbot on WhatsApp that crawls your website for live data, answers customer questions in real-time, and remembers conversations.

Workflow Overview

What is this?: This workflow deploys a sophisticated AI agent that connects directly to your WhatsApp Business account. It intelligently browses your own website in real-time to find answers to customer questions, acting as a fully automated, knowledgeable support assistant that requires no document uploads or manual training.
What problem does it solve?: Are you tired of spending hundreds of dollars monthly on AI chatbot platforms that need constant, tedious updates to their knowledge base? This workflow eliminates both the high recurring software costs and the manual labor, ensuring your customers always receive accurate, up-to-date information directly from your site.
What will you get?: You will get a fully autonomous 24/7 customer support agent on WhatsApp that provides instant, accurate answers by using your website as a live knowledge base. This dramatically reduces support costs by over 90% compared to traditional SaaS tools and frees up your team's time. Additionally, this template can be easily resold as a white-label service to other businesses, creating a new revenue opportunity.

Apps Included

How to Use

Prerequisites

An active n8n instance.

An OpenAI API key.

A Meta for Developers account with a configured WhatsApp Business App and permanent token.

A Supabase or any other Postgres database for storing chat history.

Setup Steps

Activate the Crawler Tools

Obtain your unique membership key from the provided Gumroad link. Then, open the 'list_links' and 'get_page' HTTP Request nodes and paste this key as the value for the 'auth-token' parameter in the request body of each.

Customize the AI Agent for Your Business

Open the 'AI Agent' node. In the 'System Message' field, perform a find-and-replace for '[Company Name]' and 'https://www.your-company-url.com', updating them with your actual company name and full website homepage URL.

Connect Your Service Credentials

Authenticate your services by connecting your credentials. Link your OpenAI account in the 'OpenAI Chat Model' node, your Postgres/Supabase database in the 'Postgres Users Memory' node, and your WhatsApp account in all three WhatsApp-related nodes.

Configure the WhatsApp Webhook

Copy the webhook URL from the 'WhatsApp Trigger' node. Navigate to your app's dashboard in Meta for Developers, select 'WhatsApp' > 'Configuration', and paste this URL into the webhook callback URL field.

Activate and Test Your Workflow

Save all changes and activate the workflow using the toggle in the n8n editor. To verify everything is working, send a 'Hi' message from your personal phone to your business's WhatsApp number and check for the AI's response.

Pro Tips

Enhance the workflow by adding a fallback branch. If the AI cannot find an answer, it can automatically create a ticket in Zendesk or notify a human agent via Slack.

This workflow's logic is channel-agnostic. You can easily adapt it for other platforms like Telegram or a web chat widget by swapping the trigger and response nodes.

Offer this as a white-label service to clients like e-commerce stores or local businesses. With a setup time of under 15 minutes, you can create a valuable recurring revenue stream.

Back

Information

Published date8/11/2025
Last updated8/11/2025