top of page

AI Projects

Built when I saw a problem nobody else was going to solve.

Each project started in real life, with a question that already existed and a tool that wasn't built yet. AI shortened the distance between seeing the problem and shipping a working version. That distance is where most of my consulting work lives, too.

Math Tic-Tac Toe

Web app · Spring 2026 · Built with Google AI Studio

Multiplayer math practice kids actually want to play.

THE PROBLEM

I volunteer in my kid's first-grade classroom. One morning the teacher had them racing through a paper grid called "Tic-Tac-Toe: Missing Numbers," with equations like 7 + ___ = 13 in each square. They were doing addition voluntarily, fast, because it was a game.

The paper version had limits. Same room only. Fixed problems. Once played, you'd seen the answers.

WHAT I BUILT

A web app that takes the classroom version online without losing the magic. Players pick a grade level (K through 5), choose a color, and join with a six-digit room code from any device.

Five rounds. Fresh problems each game. Confetti on the win.

WAHT CHANGED

My first grader started doing extra math at home. Voluntarily. The hard part was never the code. It was protecting what made the paper version work in the first place: competitive practice that doesn't feel like practice.

Pollen Almanac

Interactive · Spring 2026 · Built with Claude Cowork

An interactive US pollen map with live forecasts, year-round timelines, and a profile for every plant making you sneeze.

THE PROBLEM

My son's allergies turned into an ear infection. The medical plan was clear. The daily judgment call wasn't: how much medicine does a small kid actually need today?

Weather apps gave me one number. Allergy sites gave me a region. Neither matched my actual question, which was about specific plants, in our zip code, on this morning.

WHAT I BUILT

An interactive US pollen map. Search any city for today's live forecast. Browse the most common allergy-causing plants across trees, grasses, weeds, and flowers, each with photos of the plant and its pollen.

A year-round timeline shows when each one peaks. Background reading is curated from AAFA, ACAAI, CDC, Mayo Clinic, and peer-reviewed research.

WAHT CHANGED

I thought I needed a daily forecast. I actually needed an almanac. The names of the plants, the rhythm of the year, the pattern that was already there if anyone bothered to assemble it.

AI's value for non-experts isn't doing the science. It's helping you figure out what you actually need, then pulling the pieces together.

Spotting your own version of this?

Each of these started with a question that already existed and a tool that wasn't built yet. If you're sitting on a similar gap inside your team, I'd love to hear about it.

bottom of page