In plain terms: this turns your documents into a smart Q&A brain. It reads and remembers your content, then answers questions accurately using WhatsApp, OpenAI (LLM), OpenAI Embeddings — great for support and internal knowledge.
18
Nodes
4
Integrations
16
Connections
WhatsApp, Manual
Trigger
🚀 Deploy this workflow
Deploy this automation in minutes
Log in and we'll walk you through every key step by step, then deploy it for you.
The real node graph exported from n8n — triggers (green), AI nodes (violet), services and logic — laid out as they run.
⚙️ How it works
WhatsApp Trigger· whatsAppTrigger
OpenAI Chat Model· lmChatOpenAi
Window Buffer Memory· memoryBufferWindow
Embeddings OpenAI· embeddingsOpenAi
OpenAI Chat Model1· lmChatOpenAi
When clicking ‘Test workflow’· manualTrigger
Embeddings OpenAI1· embeddingsOpenAi
Recursive Character Text Splitter· textSplitterRecursiveCharacterTextSplitter
Handle Message Types· switch
AI Sales Agent· agent
Product Catalogue· vectorStoreInMemory
Vector Store Tool· toolVectorStore
get Product Brochure· httpRequest
Create Product Catalogue· vectorStoreInMemory
💼 Business analysis
🎯
Business use case
Builds a retrieval-augmented knowledge system: content is embedded into a vector store and served back as grounded, citeable answers. Ideal for support deflection, internal knowledge bases, and AI assistants that must stay factual.
📈
Market need
Generic chatbots hallucinate; companies need answers grounded in their own documents. RAG is the proven pattern, and a working pipeline is scarce and valuable.
💰
Market value
RAG assistant implementations typically run $5,000–$20,000 bespoke. Packaged, it underpins support-deflection tools that sell for $99–$999/mo.
🧩 Integrations & stack
🟢WhatsApp🧠OpenAI (LLM)🧬OpenAI Embeddings🌐HTTP / REST API