Mechanical + EECS · MIT ’27
AJ Santa Cruz
I build precision hardware — and the simulation and ML that make it work before it exists.
I work across the full stack of a physical system: mechanical design and FEA, custom motor-control electronics, and the machine learning that turns noisy sensor data into decisions.
View the workApproach
De-risk the hardware before it’s built.
Models precise enough to trust, and simulation that kills a wrong idea in an afternoon — so the physical build is the last risk, not the first.
- Mechanical & simulation
Static, modal, and harmonic FEA; multiphase CFD and thermal — run to make a decision, not a picture.
- Electronics & controls
Custom motor-control hardware — field-oriented control, back-EMF sensing, thermal management — on multilayer boards.
- Perception & ML
Signal-only, CPU-only models that survive the gap between clean lab data and a real, noisy machine.
Selected work
Thirteen projects across perception, hardware, and simulation.See all →
Real-Time Machining-Chatter Detection
~97% accuracy on noisy real-world signals, trained only on clean data, where standard models collapse to ~46%.
View projectPrecision-Motion Digital Twin
Sub-25-micron accuracy under load, and a failed 100 GB simulation rebuilt into a 34-second solve.
View projectAutonomous Tissue-Targeting Insertion Device
A loading mechanism that unlocked 5x sample capacity on standard plates with zero new consumables.
View project
360° Thermal-Imaging Drone for Crack Detection
A custom, government-compliant autonomous drone reaching 55 mph, with electronics and structure I designed and FEA-validated.
View project