Where this page stands
Short version only, written from my resume notes. The projects below were internal, so anything I add later has to stay at the level of what the work was and why it mattered, not how the systems are built. Expanding this is on my list.
Overview
I spent summer 2026 in fleet performance management on the onshore wind side of GE Vernova. The role was undefined when I arrived, so I scoped it myself: I went to platform, controls, and reliability engineers, asked what was actually slowing them down, and turned that into five projects I could finish before August.
What I worked on
Blade serial numbers out of drone imagery
A Python tool that pulls drone inspection imagery, reads the blade serial number off it with a vision model, and reconciles what it finds against fleet records. The inspection vendor had quoted fifty thousand dollars for the same capability.
Controller fault triage
Automated turbine controller fault triage in Python and SQL, routing qualifying flags to controls or platform engineering. Saved three to six engineer hours a week.
Measurement system analysis
Ran an MSA across two fleet databases to settle which source forced controller variables should be measured from, and corrected 54 erroneous turbine configuration records along the way.
Still to write
- What fleet performance management actually does, for people outside wind.
- The other two projects.
- What I learned about how a mature organization treats failure, which is the part I actually think about.
- A cover image that is mine and not company branding.