Handle WhatsApp sales queries with GPT-4, Supabase, and a product catalog

Go to Workflow
3 views
Built by Shadrack Shadrack
Created on June 05, 2026

Description

How it works
An AI-powered sales agent on WhatsApp that handles product inquiries using your Supabase knowledge base and n8n catalog. Customers can send text, voice notes, or images to ask about products, pricing, and specs. The agent responds with detailed answers, product images, and FAQs, creating a complete self-service sales experience.
(I primarily designed this for furniture business, consider tailoring it)
Setup steps

Connect Supabase credentials for your knowledge base and FAQs
Configure n8n tables with your product catalog (prices, descriptions, image links)
Set up WhatsApp Business API integration
Add your product categories and common queries to the AI context
Test with sample product questions and image uploads

Customization tips

Structure your catalog tables with clear columns (SKU, price, description, image_url)
Add industry-specific terminology to the AI prompt
Create templated responses for common FAQs to ensure consistency
Enable voice-to-text transcription for better voice note handling
Add product recommendation logic based on customer queries
Set up image recognition for customers sending product photos to identify items

Nodes Used (12)

AI Agent
@n8n/n8n-nodes-langchain.agent
Code
n8n-nodes-base.code
Default Data Loader
@n8n/n8n-nodes-langchain.documentDefaultDataLoader
Embeddings OpenAI
@n8n/n8n-nodes-langchain.embeddingsOpenAi
HTTP Request
n8n-nodes-base.httpRequest
OpenAI
@n8n/n8n-nodes-langchain.openAi
OpenAI Chat Model
@n8n/n8n-nodes-langchain.lmChatOpenAi
Recursive Character Text Splitter
@n8n/n8n-nodes-langchain.textSplitterRecursiveCharacterTextSplitter
SerpApi (Google Search)
@n8n/n8n-nodes-langchain.toolSerpApi
Simple Memory
@n8n/n8n-nodes-langchain.memoryBufferWindow
Supabase Vector Store
@n8n/n8n-nodes-langchain.vectorStoreSupabase
WhatsApp Business Cloud
n8n-nodes-base.whatsApp