About
The same skill, at a different layer. A filament-feeder firmware bridge that replaces a proprietary RS485 link with plain USB-C UART, and a laser-diode projector rig, sit next to machine learning that reads a VIN plate or an odometer straight from a photo and a repair bench that diagnoses the fault before it's quoted. Four different surfaces, the same instinct underneath: work out exactly what's happening before touching anything.
Most technical businesses land on one side of that line and stay there. Few treat a stuck 3D-printer control board and a stack of insurance-claim photos as the same kind of problem — work out what's actually happening, then build the fix.
What spans both
- Firmware & embedded
- Control software for hardware that doesn't already speak the language it needs to — protocol bridges, not just app code.
- Signal-level diagnostics
- A fault traced to the exact component on the bench, not swapped-and-hoped.
- Applied machine learning
- Computer vision and document extraction that reads a photograph well enough to replace typing the number in by hand.
- Acoustics & audio systems
- Room measurement, system design, and the signal chain that gets a PA behaving.
No portrait published yet.