Bio
Engineer working at the intersection of mechatronics, embedded systems, and applied machine learning — pursuing the question of how a system, given good feedback and a few honest assumptions, learns to behave well over time. Placeholder bio — will be replaced once interview block 1 is answered.
A letter
This site exists because somewhere between every clean PDF on a hard drive and the conversations recruiters and collaborators actually have, an enormous amount of context goes missing. The numerical methods that took two weeks of thinking, the embedded firmware that crashed on the demo stage and was fixed in twenty minutes flat, the moment a control loop finally settled — none of that lives in a transcript. So the goal here is simple: surface the thinking. The interactive demos are the centrepiece because they are the only way to honestly say I learned this; here, you can play with it too.
This is, secondarily, a love letter to the courses themselves and to the people who taught them. If a younger engineer lands here and decides to spend an extra weekend on Fourier transforms because the epicycle animation was beautiful, that is a win.
Placeholder letter — will be replaced once interview block 2 is answered.
The journey
How it started
The spark was somewhere between an early childhood obsession with how things fit together and the first time a piece of code I wrote made a physical thing happen. (Specific moment will replace this sentence after interview question J1 is answered.)
What broke me, and what fixed me
Engineering education does not move in a straight line. There were courses that genuinely shook the framework I had brought in — usually around the point that the math stopped being mechanical and started demanding intuition. (Specific course/project to replace after J2 + J3.)
The clicks
Some understanding arrives all at once. The moment a Bode plot stopped being decoration and became a story; the moment “convolution” stopped being a symbol on a page and became a sliding-and-multiplying gesture; the moment a PID loop stopped feeling like a recipe and started feeling like a conversation between three opinions about the future. (Specific ‘aha’ moments to replace after J4.)
Failures I keep
(To replace after J5.) Some mistakes I am genuinely glad I made — they taught the kind of lesson that no amount of reading would have produced.
Where I am headed
(To replace after J6.) The work that interests me most is at the seam between three things: the math that lets a system reason about its own state, the firmware that runs in milliseconds, and the design conversations that decide which problem is worth solving in the first place.
Advice for the version of me starting out
(To replace after J7.) Spend more time being wrong on purpose. The cost of an idea you tested and discarded is far lower than the cost of the habit of not testing.
Get in touch
Built with Astro, Claude, and a lot of old PDFs.