North Sioux City, South Dakota · Class of 2027
I won a state championship on a mountain bike. Then I rode 150 miles over twelve hours to see whether I could. After that, I started building things, each one harder than the last. At my internship, I built a system that determines where new inventory should go in a warehouse to reduce the cost of later picks and put-aways. I rebuilt Catan in code and simulated nearly a million games to understand how its economy works. I built Vale, a styling app, to simplify a decision I make every day. Now I am building a simulation of how a person thinks and decides. What happens independently, and what reaches awareness? The model separates a body that acts on its own, a thought stream that is often elsewhere, and a gate that determines what enters awareness.
Now
Research · 2026 – present
A simulated human made of two systems: a body that acts automatically every tick, and a language-model thought stream that is usually thinking about something else. An awareness gate decides what reaches thought. Change one automatic parameter, rerun the same day hundreds of times, and measure what changes.
Company · 2026 – present
Dress well. Don't think about it. Vale learns what you own, reads what your day asks of you, and puts one outfit in front of you each morning. It is a decision-elimination app, not a fashion app
Simulation · 2025 – 2026
A rules-exact simulation of the board game, six styles of bot, and about 987,000 simulated games, read two ways. As an economy: ore and wheat trade at a premium because the recipes use them up, the bank caps every price, and a player who controls a good sells it cheap. As a strategy problem: where you settle decides more than how you play.
Internship · Tegra Corporation · January 2026 – present
As a data analytics intern I built a system that decides where a new item should be placed in the warehouse so the whole operation runs most efficiently. Before that I pulled the company's invoice history with SQL, lined it up against Federal Reserve data on machinery orders, and found a four-month lead, then showed that three customers explain most of the swing in revenue. The dashboards that came out of it are the ones the business reads.
Also built
Each solved a problem I actually had.
I tutor students for the ACT, so I wrote the software I run it on: a student side with on-screen and paper diagnostics, a tutor side with assignments, notes and answer review, and server-side scoring from the book's raw-to-scale tables. Built with React, Supabase and Cloudflare.
Stats