I build predictive decision tools for weather-exposed operations.

For logistics, freight, and distribution teams whose demand moves with the forecast — and who are still calling it from a spreadsheet.

Book a 30-minute call

Two ways to start

Forecast Pilot2 weeks: We take one recurring decision your team makes ahead of the weather — crew sizing, load timing, dispatch windows, stocking ahead of a season — and I ship a working forecasting tool against your own historical data. You keep the tool and its source whether or not we continue.
Manual Process Teardown3 weeks: There is an analysis someone on your team runs by hand every week, because automating it has never quite been worth the trouble. I automate it and hand back the tests that prove it stays correct. At the Air Force Weather Agency I replaced a 40-hour manual validation of 16,000 daily products with a suite that ran in 30 minutes.

I was the forecaster before I was the engineer.

B.S. Meteorology, minor in Mathematics: University of Nebraska, full scholarship
U.S. Air Force weather officer: 2000–2008 · nine years of operational forecasting for DoD training and missions
Veterinary pharmaceutical purchasing: 2008–2015 · $5M+ annual spend · 10,000 SKUs · 100 vendors · forecast by hand
Air Force Weather Agency, via Northrop Grumman: 2016–2020 · validated ~16,000 environmental products daily across 150,000+ data points; a 40-hour manual check became a suite that finished in 30 minutes
U.S. Strategic Command, via Northrop Grumman: 2020–2022 · geospatial displays commanders used for situational awareness
Operations software, RTG Medical: 2022–2024 · hiring, payroll and placement across 1,000+ clinicians and 100+ facilities

Also built

Analog Jones and the Temple of Film
A film podcast's public site and internal admin, built as an Angular 21 monorepo on Firebase and Cloud Run, with an AI pipeline that turns each episode’s audio into its written writeup.

How I got here

What I did

I spent nine years as an Air Force weather officer, then left active duty in December 2008 and went to work at a buying desk.

For the next six years I bought veterinary pharmaceuticals — roughly 10,000 SKUs across 100 vendors, north of $5M a year. Flea and tick medication, wormers, West Nile vaccines. Demand for all of it moved with the weather, one to two months ahead of the season, and I forecast it in a spreadsheet: last year's sales, new products entering the market, and my own research into which regions were heading for a heavy season.

Getting it right was never really about inventory. It was about money. A large, well-timed seasonal order bought NET 90 terms instead of NET 30, plus a bulk discount, so the product sold before the invoice came due. The forecast paid for the float.

What I built

I learned to build software because the buying desk needed it. Pay-per-click in 2009, out of curiosity and necessity. HTML, CSS, and JavaScript in 2012, to rebuild the company's storefront. I never stopped buying — I just kept adding the tools I wished existed. $2M in annual online sales growth, organic traffic up 50%.

Since 2016 I have built it for other people making that kind of call. At Northrop Grumman, working at the Air Force Weather Agency, I validated the environmental products operational forecasts are built on — about 16,000 a day, 150,000 data points — and replaced a 40-hour manual check with a test suite that finished in 30 minutes. Then at U.S. Strategic Command I built the geospatial displays commanders used for situational awareness. After that, four years of operations software: the systems a staffing company ran hiring, payroll and placement on, for a thousand clinicians across a hundred facilities.

What I wanted back at that desk was research that ran itself. That is what WillaWave builds.

Get in touch

A 30-minute call is the fastest way to find out whether this is worth doing. I read every message myself and reply within one business day.