I didn't set out to build an AI company.

I co-founded BetterDog, an in-home dog training business. Like any small business, ours runs on bookings, calendars, payments, memberships, notes, and follow-ups, each living in a different tool.

So I started solving the small, practical problems: capturing what happens in a session, keeping a knowledge base, letting helpers handle the repetitive work. I wasn't trying to build a theory. I was trying to get my time back.

The 4 a.m. realization

One morning it clicked. What we'd been building wasn't a distraction from the business. It was a working prototype of something much bigger: an AI that understands a person, navigates the systems for them, and keeps them in control.

BetterDog already had the ingredients of a real test: real clients, real sessions, real payments, and real consequences when something goes wrong. That's rare, and it's why I think the lessons here are worth sharing.

Why it matters to me

People and small businesses are drowning in software. Every new tool promises to save time and then asks you to learn it, log into it, and maintain it. The answer isn't one more app. It's one advocate who works across all of them.

That's what I'm building, in public, starting with a dog training business. Here's what that looks like so far.