An inbox with direction / Selected work
AI Email Triage Automation
n8n-based workflow that receives inbound email, sends it through a local Node bridge, and returns structured action recommendations.
Prototype01 / The project
Behind the work.
This automation project combines n8n workflows with a local HTTP service that calls an LLM CLI to classify incoming mail. It returns structured fields such as importance, action requirements, summary, suggested next steps, and deadline mentions. The project is useful because it shows practical integration work: workflow orchestration, a local service boundary, and a structured output contract for downstream automation.
Turn an inbox into next steps.
Explore three example recommendations.
Incoming sample
Please review the proposal by Friday
Could you review the attached proposal and send your comments before Friday?
Structured recommendation
Action requestedReview the proposal
Friday (as written)
Prewritten examples of structured triage output, not live AI classification.
03 / The build
Details that matter.
Uses n8n workflows to receive and route messages into a local processing service.
Returns structured fields instead of raw text so the output can feed follow-on automation.
Shows pragmatic AI integration through a clear service boundary rather than bolting prompts into UI code.
Workflow Layer
n8n / IMAP Trigger / Workflow JSON / Smoke-Test Workflow
Bridge Service
Node.js / HTTP Endpoint / CLI Invocation / Structured JSON Output
AI Output
Importance Scoring / Needs-Action Flags / Summaries / Suggested Actions
Have a related challenge?
Let’s talk about your project.
Tell me what you are trying to build, who it is for, and what a useful result would look like.