A few months ago I opened Claude for the first time. I felt stupid.
For context: I taught myself C++ from a book in middle school. C++ for Dummies. I thought I was such a cool kid (just kidding, I knew I was a nerd).
Then I went to high school, made some friends, joined the family business, and didn't touch code for 20 years.
So when I opened Claude, I had just enough background to know how much I didn't know. Which might be the worst place to start.
I kept going anyway. Here's why.
Real estate has been late to every technology wave. I didn't want to be the operator who waited again.
The pushback was immediate.
"Early adoption won't move the needle."
"You're a real estate shop, not a tech company."
Here's what I actually think is happening when you learn to use AI:
The tool isn't the point. The rewiring is.
I used to evaluate a deal by asking: does this work? Now I ask: under what conditions does this break? It allows us to be much more specific on downside protection and that’s important to me because my family puts in at least 10% of the equity on every deal we do.
It's forcing me to rethink every part of our workflow — what deserves my attention, where the real bottlenecks are, what I've been doing manually out of habit for years.
Here's a small example. We built a Python integration so Claude can run sensitivity analyses directly in Excel. Tweak debt assumptions, exit caps, rent, downtime, lease-up pace — across dozens of scenarios at once.
We're not underwriting faster.
We're underwriting wider. Deals that used to take a week to evaluate now take a day. That's a bigger net. That's more deal flow.
Real estate has always rewarded operators who could see more deals and underwrite them more accurately.
AI doesn't change the game. It changes who can play it at scale.
I still feel uncomfortable every time I build something new.
That's probably the signal I'm doing it right.
Are you building your own tools — or waiting for someone to hand you the perfect one?
Discussion about this post
No posts
