Now

I’m spending most of my time on robotics engineering work that sits between research and real-world use. The day-to-day work is a mix of software integration, perception, system reliability, testing, and making complex robot systems easier to understand and operate.

I’m also working on language-model interpretability, mostly around understanding how model behavior emerges across intermediate layers. That gives me a second thread alongside robotics: one closer to model internals, representation analysis, and reasoning about what a model appears to know before its final output.

Across both threads, I’m interested in making complex systems less opaque, less fragile, and easier to work with.

At the moment, I’m reading:

• Replaceable You by Mary Roach

For downtime, I’ve been exploring Abu Dhabi, reading, and playing Apex Legends.

(This is a now page, and if you have your own site, you should make one, too.)