Mechanical engineer at Apptronik working on humanoid robots. Started in pure CAD; the work has pulled me deeper into simulation, scripting, and the small custom tools that let a hardware team move faster. The interesting problems live where the hardware meets the code.
Working on humanoid robots — the day-to-day blends mechanical design with simulation, scripting, and custom tooling that helps the hardware team iterate faster. Promoted IC1 → IC2 → IC3 in two years.
A mechanical engineer building the bridge between hardware and the software that drives it.
• Simulation data pipeline — Python tooling that parses time-series logs from policy executions into actionable load cases and metrics, closing the loop between simulation and hardware design.
• Architecture for a next-gen platform — driving the foundational layout. Writing journal scripts to bridge simulation outputs into engineering-ready master models in CAD; the API is sparsely documented, so I lean on agentic LLM workflows to get reliable generated code.
• Earlier in tenure — led an arm subsystem to CDR; resolved a critical thermal issue with a hybrid metal / 3D-printed redesign that doubled as a structural fix. Designs from that phase ship at fleet scale today.
• URDF ownership — saw a hardware/software handoff gap and proactively took over the URDF generation and post-processing pipeline, becoming the bridge between the two divisions.
Open to chats with humanoid teams hiring on either side of that line.
A working set, not a resume bullet list. The things I open every week.
Earlier work in vehicle dynamics, transmission design, and product development. Recent humanoid robotics work is under NDA — happy to talk about it in conversation.
Standing up an RL locomotion pipeline for a humanoid on my own box — and asking the mechanical-engineering question of which physics-step data actually informs hardware decisions.
Two small Node services I wrote so my agent fleet is reachable from my phone — one fast opinionated relay, one scoping companion with persistent memory. Running daily.
Complete redesign addressing bearing wear, oil leak, and weight on the previous Terps Racing car. New belt path enables AWD output.
Successor to TR22 — interfaces to a new 4WD clutch, supports updated loadings, tightens manufacturing tolerances based on field data.
Ground-up rework of a 2018-era steering rack to reduce driver effort, raise yield strength, drop weight, and make in-pit assembly faster.
Designed and built in 4 months on a $500 budget — a moving rig that cleans large-format solar arrays without water. Concept to functional prototype.
First serious CAD work — turret and intake mechanisms for the team's competition robot. The project that committed me to engineering.
Took on the CAD-lead role; focus shifted from solo modeling to scaffolding sketches and standards so the rest of the team could move faster.