automate3–4 hoursIntermediate

Build an AI That Triages Your Inbox

Maps to: AI Application Builder · Executive Assistant · Chief of Staff · Operations Analyst · Solutions Engineer

You're going to build an AI agent that reads your incoming email, sorts it into categories, and drafts replies for the ones that need them. The skill is agent design: deciding the categories, then drawing the line on what the agent handles on its own versus what it must flag for you. That's the textbook first AI agent and a core thing AI application builders ship, and doing one tells you fast whether designing what an agent is allowed to do is your kind of work.

How this shows up on a resume or college app

I built an AI agent that triages my real inbox, designing the categories, fixing its misclassifications, and setting deliberate guardrails on what it can handle on its own vs. escalate. I learned that AI agents work best when paired with clear rules and boundaries, not asked to figure it all out.

When you finish, BuildMe drafts your Common App activity description from what you actually built.

Start this project

The plan

1
Step 1
Map your inbox categories
Decide the categories your agent will sort into, and which one matters most to get right. This taxonomy is the agent's whole brain; vague categories = a useless agent.
2
Step 2
Build the triage logic + get it sorting once
Build the agent: a trigger on incoming mail → the AI classifies → a branch per category → draft a reply where needed. Run it on a few emails so you've seen it sort.
3
Step 3
Run on real mail + fix misclassifications + set guardrails
Point it at your real inbox. Your clean categories will fall apart on real mail, so expect to redo the rules a couple times; that's normal. Fix the misclassifications, and decide the guardrails: what must it NEVER auto-handle (e.g. never auto-reply to a teacher or a stranger)?
4
Step 4
Run for 3 days + close
Let it run on your real mail for 3 days and refine. An agent you trust enough to keep running is the proof.