About

Hey, I'm Brad. I've been writing software professionally for over 15 years. I started in computer graphics and have since bounced between rendering, distributed systems, platform engineering, and ML - but the common thread is caring about performance, reliability, and developer experience.

I started out building real-time rendering for flight simulators at Diamond Visionics, which led me to grad school at Binghamton where I researched software ray tracing - figuring out how to visualize massive scientific datasets on CPUs when GPUs weren't available. During my master's I interned at Intel working on the OSPRay ray tracing framework. After graduating I wrote sensor fusion software for Navy helicopters at Lockheed Martin, then went back to Intel to optimize Metal and OpenGL drivers for macOS. Through all of that I was also working towards a PhD in natural language processing before deciding that it just wasn't something I could manage while working full time.

I spent five years at Google working on Stadia and related projects, where I built the Vulkan video pipeline, developed features like StreamConnect, and eventually helped integrate Stadia's streaming tech into Android XR. After Google, I joined Daydream as an early engineer, architecting the backend platform and designing an LLM-powered semantic search system. Most recently I led platform engineering at Digg, focusing on operations, reliability, and developer experience.

Outside of work, my hobby is collecting new hobbies. This site is where I write about whatever I'm building or learning at the moment.