I'm Alex — an independent embedded systems developer, and the one and only employee of Rosito Bench.
I grew up watching the birth of the machines I still build for. My first gadget was a TI Dataman, its vacuum-fluorescent display glowing that particular blue-green only 1970s tech seemed to know how to make. Then a VIC-20, then a TI-99/4A, where I actually learned to program. I once owned an Osborne 1 — advertised as "portable," which in 1981 meant "has a handle" and weighed as much as a car battery. My cousins had a Commodore 64 and, in one lucky house, a Macintosh and an NES. After school, it was the arcade — Space Invaders first, then Pac-Man and Donkey Kong arrived and I lost my mind a little. That was the same moment Zilog's chips and their contemporaries were quietly starting the revolution I'd spend my life inside of. I learned CP/M before I learned DOS, and never really stopped tinkering since.
I never had a formal engineering degree. What I have instead is four decades of taking things apart, a background teaching database design, and a stubborn habit of solving problems the way I have the tools to solve them — which usually means: small, direct, and without dragging in more than the job requires.
Rosito Bench exists because I kept building the same kind of tool for myself, over and over, and eventually noticed the pattern. A thermometer I built four years ago — because my wife asked what temperature it was in the room, and I thought, I can do that — is still running today, recalibrated more than once, still telling her the answer. Hasaki, my first and most personal project, trains neural networks on a desktop and hands you back a single C header with no runtime attached, built for the microcontrollers everyone else forgot about — free for personal use, with a Pro edition for anyone building on it commercially. The rest of the bench grew from the same instinct: build the small thing that's actually needed, skip everything that isn't.
I work from a 1990s glass desk that's far too narrow for the job, surrounded by instruments with more history than most of my code: a Simpson 260 multimeter I found at a Goodwill for ten dollars, a Radio Shack meter that belonged to a friend who isn't here anymore, an NRI Discovery Lab my wife has banished to live under the couch. None of it is fancy. All of it still works.
That's the whole philosophy, really. Tiny tools. Big solutions. Nothing you don't need.