- NOW
- Intern, Daimler Truck North America
- EDUCATION
- Oregon State: M.S. Computer Science (current), Honors B.S., 3.95 GPA
- PREVIOUS
- Software Engineering Intern, DZYNE Technologies
- DESIGN
- Photography, and the design of this site
Hey there! My name's Henry. I'm a Computer Science master's student at Oregon State University, where I also finished my honors undergrad, and I'm joining Daimler Truck North America as an intern. I'm passionate about machine learning, space, and medicine, and I love working on software and impactful technology that helps people. I'm driven by problems where computation meets real-world change and improvement. Outside of engineering, I'm usually behind a camera, on the slopes, lifting, or finding new music. I value growth and learning above everything, and I'm always excited to connect with others who share that mindset!
Daimler Truck North America
Commercial Vehicles
I'm joining Daimler Truck North America, the company behind Freightliner, as an intern. It's early, so this entry is short on purpose: I'll write it up properly once there's work here I can talk about.
DZYNE Technologies
Embedded Systems & Full-Stack
I worked on a small team writing embedded C and C++ for anti-drone defense systems. The codebase had grown organically and needed serious cleanup: I refactored the core modules to be properly modular, which made a real difference in how fast the team could move. I also rebuilt the Python test infrastructure from scratch because the old one was slow and required too much manual babysitting. The most fun part was building a full-stack GUI in React and Flask that gave operators real-time control over power, tracking, and logging during test runs: something that previously meant running a bunch of scripts by hand.
Plasma, Energy, and Space Propulsion Laboratory
Signal Processing & ML
I split my time between the thruster side and the biomedical side of the lab, doing ML and signal processing work on both. A big chunk of it was building pipelines to extract clean signals from really noisy sensor data: plasma environments are brutal for that. On the modeling side, I worked on predictive models trained on large experimental datasets. One of the more interesting problems was automating capacitor tuning for RF plasma systems using Google OR-tools, replacing a slow manual process with something that ran dynamically and maximized power coupling in real time.
Jason Clark Research Group
FPGA & DSP Engineering
The lab works on precision sensing at the micro and nano scale, and my job was building the FPGA-based DSP system that let them actually measure the signals they cared about. Nano-ampere acquisition was something the lab hadn't been able to do before, and getting there meant writing VHDL modules, building thorough testbenches, and then integrating artificial damping algorithms through Hardware-in-the-Loop testing with Moku instrumentation to get the sensors stable enough to trust.
- OPT-01Cap_
Match_ NetPicks capacitor values for an RF matching network from the parts the lab actually stocks.Python · OR-Tools CP-SATQUALIFIED100% - SYS-01small-shellA Unix shell in C with job control, I/O redirection and signal handling.C · POSIXQUALIFIED100%
- MDL-02floralyticsRanks the plants worth planting in a county from the Oregon Bee Atlas.Python · FastAPI · ReactQUALIFIED100%
- APP-03accliMateAnswers questions about an unfamiliar codebase and cites the lines behind each answer.Python · tree-sitter · ChromaDBQUALIFIED100%
- APP-01animAgentShows running Claude Code agents as characters in a pixel-art room.Swift · SpriteKit · AppKitQUALIFIED100%
- TUL-02groupStatPRIVATERanks a group chat eight ways from the tapbacks in a macOS Messages database.Python stdlib · WKWebViewQUALIFIED100%
- TUL-03Ground-ControlEvery repository on my disk with its git state, running agents and onboarding doc.Node stdlib · SwiftFLIGHT90%
- TUL-04orchestratePRIVATERuns a team of Claude Code agents at a project as a dependency graph.Node stdlibFLIGHT90%
- APP-02GrowthAppPRIVATEA habit tracker that draws each streak as a filament of one graph.SwiftUI · SwiftData · WidgetKitFLIGHT62%
- QNT-01aggregateAnalyticsA margin-aware Elo model that predicts a margin for every NFL game.Python · polars · scikit-learnFLIGHT60%
- QNT-02ML_
quantitative_ researchMonte Carlo portfolio risk: the distribution of outcomes, especially the tail.Python · NumPy · SciPyFLIGHT70% - MDL-03gpt-scratchTransformer internals written by hand, from a single neuron up toward a GPT.Python · PyTorchBREADBOARD40%
- MDL-01emnistA feed-forward network with no framework, trained on EMNIST characters.Python · NumPyQUALIFIED100%