Clone yourself: turning your job description into a personal AI agent
The four tasks I'm automating now
Showing feedback. After every showing with a buyer, the listing agent gets a note — what the client liked, what didn't land, whether there's a second look coming.
The voice memo happens in the driveway on the way out, usually one minute, basically a brain-dump while the showing is still in the air. The AI agent drafts the email from that. I edit, send.
The voice is mine because the AI agent was trained against a stack of feedback emails I'd already written by hand — that's the corpus the voice anchors to. Time per: down from 10 minutes to two.
Overnight changes. Every morning, a one-page summary of what shifted on the MLS in the markets I'm watching — new listings, price changes, status flips, mortgage interest rates, the listings that went pending and the ones that came back.
The AI agent pulls it before I'm awake. The point isn't speed. It's that I actually read it now, in a format and with visuals I like, where before I'd skim various sources and miss things.
AI Generated Home Pulls. Buyer says: "three-bed, 2.5 bath, under 850k, walkable, prefers older homes with character, positioned East of their office so they don't drive into the sun on the way home and on the back, and within 12 minutes driving distance to their work."
The AI agent runs a daily scheduled automation every day after 2:30pm — when MLS submissions are closed. It checks against active inventory using Property Finder (the tool I built two months ago and have given constant daily feedback to ever since) and returns a shortlist with the reasoning attached. It also has knowledge of what the buyer liked and didn't like about previous showings.
For this particular buyer, we'd had zero results every day because of how specific the criteria is. Today was the day it turned up a property.
I review everything it says, double-check the status on the MLS, send it off to the client. They love it. We schedule a showing. We laughed together as every box was checked off the list and then some.
Monday SOI prep. Every Monday morning, Tahsin and I sit down with a Sphere of Influence Playbook — a one-pager covering our active buyers, sellers, hot / warm / cold leads, upcoming home-iversaries, last-interaction recency, wins, losses, and the to-dos that surface from all of it.
The whole thing runs out of an internal AI tool I built to keep the SOI data organized and analyze it weekly. The AI agent assembles the one-pager Sunday night.
The thinking is ours. The setup is the AI agent's.
The point isn't the one-pager. It's that we both walk in already having the same data in our heads, so the hour gets spent on action items and strategy instead of worrying about what slipped through the cracks.
All of this might seem out of reach. A few weeks learning technologies like Claude Cowork is what it took.
What stays human
Put together a list of everything you do on the computer. Presently, with AI technology you can automate or guide the AI to do almost all of it. By taking the time to do this you open up your time to focus on what only humans can do — at least until we have affordable AI humanoids everywhere.
The offer
If you want to catch up on the ways AI can solve problems and eliminate the tedious, time-consuming tasks in your personal or professional life, feel free to send me a list of what you're dealing with. I'll send back what I'd hand off to AI first. Then if you need help setting up a personal AI dashboard and getting trained on it, I'd be happy to help.
