I'm a software engineer in St. Louis working on machine learning: how fast models really run, why, and the systems they run on. By day I secure Kubernetes and Azure delivery at Boeing. Alongside it, a document platform nobody asked me to build, and a master's in data science at the University of Illinois.
Outside of that: scuba diving, rock climbing, hiking, and making things that aren't software. I like problems where the consequences are real.
Open to consulting and interesting problems
5+ yrs
Shipping mission-critical software
$90M
Contract award my cloud migration demo helped secure
10×
Faster integration test cycle
4×
Fewer engineering hours on porting
How I can help
Everything below is work I do now, not work I did once. If your problem looks like one of these, we should talk.
Classifiers and NLP pipelines that ship — data preparation through evaluation through deployment — and the performance work to know what they cost to run across PyTorch, JAX and TensorFlow. I care more about the failure modes than the leaderboard.
Threat modeling, zero-trust delivery pipelines, service-mesh networking, and security review built into the release process instead of bolted on at the end. This is my day job.
Platform engineering, Kubernetes operations, cloud-native systems, and the boring-but-critical details of reliability, automation, and production support. If it has to run at scale without surprise failures, this is the work I do daily.
Build systems, CI pipelines, test harnesses and internal tools. Most teams are one good tool away from moving twice as fast, and nobody has time to build it.
Selected work
Some of it shipped to production on hardware I'll never see again. Some of it is a side project my wife asked for. Both taught me something.
A framework-neutral Python library that tells you how fast a neural network should run, not just how fast it does — FLOPs, tensor traffic and a roofline model across PyTorch, JAX and TensorFlow.
Control any smart device with American Sign Language — hand tracking, a gesture classifier and grammar correction, wired together over a client-server link.
An NLP pipeline that routes patient portal messages to the right team and surfaces relevant evidence-based research while they wait for a human.
Writing
Years after our college capstone, an out-of-the-blue email from European AI researchers wanting to reproduce our sign language paper sent me to Google Scholar, where we've been cited 11 times.
Looking back at my senior capstone project at Missouri State — building a real-time American Sign Language smart home controller with MediaPipe, custom CNNs, and networking — and how it sparked my path in machine learning.
The neural-cost v0.2.0 GPU benchmark moves from CPU roofline analysis to Apple Silicon's MPS GPU backend — measuring PyTorch and JAX across five architectures, four batch sizes, and both eager and compiled execution modes.
Off the clock
Same habits, different problems. Most of what I'm good at at work, I practised somewhere else first.
Underwater
Checklists, gas planning, and a buddy who checks your work. Turns out I like that underwater too.
On the wall
Problems you solve with your hands, where the feedback loop is very short and very honest.
Player two
Research, I tell myself. One of them ended up as a reinforcement learning project.
Tetris RLSketchbook
Things that aren’t software, with nothing to compile and no tests to pass.
Let's connect
Contract work, a full-time role, or a problem you want a second opinion on — I read every message and reply to the ones that aren't a template.
Usually replies within a day or two · St. Louis, Missouri