In plain terms: this is an AI assistant that works the moment a new message arrives. It uses AI to understand and generate content, so it handles judgement-style work, not just moving data. It connects to OpenAI Embeddings, Gmail, Supabase Vector, OpenRouter (LLM) and completes multi-step tasks for you automatically — the kind of work you'd otherwise hand to a virtual assistant.
29
Nodes
6
Integrations
37
Connections
Gmail, Chat
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
Embeddings OpenAI· embeddingsOpenAi
Chat Memory· memoryPostgresChat
Output Parser· outputParserStructured
LLM· lmChatOpenRouter
incoming_email· gmailTrigger
AI· lmChatOpenRouter
When chat message received· chatTrigger
retrieve CONTEXT· vectorStoreSupabase
Kate· agent
Get a message· gmail
Email Labeler· chainLlm
RAW MIME for draft (threaded)· code
clean text· code
LABEL· set
💼 Business analysis
🎯
Business use case
Deploys an autonomous AI agent that can reason over instructions, call tools, and take multi-step actions on behalf of a team — replacing repetitive knowledge work. It orchestrates OpenAI Embeddings, Gmail, Supabase Vector, OpenRouter (LLM), HTTP / REST API, Airtable so a single natural-language request becomes a completed task.
📈
Market need
Businesses are racing to put AI agents to work but lack the engineering to wire models to their real tools and data. This delivers that wiring as a ready-to-run asset.
💰
Market value
Custom AI-agent builds are quoted at $3,000–$15,000+ by agencies, with managed retainers on top. As a productised template it commands premium one-time or subscription pricing.