AI engineer. I build machine-learning systems that ship — clinical intelligence for healthcare, generative discovery for materials science. From on-device neural inference to models that search chemical space at machine speed. This city is my work. Fly through it.
I engineer AI for healthcare — embedding-driven matching at population scale, on-device neural inference at the edge, closed-loop pipelines that turn raw clinical signal into decisions in milliseconds. Privacy-preserving by design. Sovereign by default. Built to run where the patients are.
In materials science I point deep learning at the periodic table — graph neural networks for property prediction, generative models that traverse chemical space orders of magnitude faster than any lab, simulation-in-the-loop pipelines that close the gap between hypothesis and synthesis. The search space is effectively infinite. The model doesn't blink.
Essays on humans and AI, written before it was a headline — where the machines are taking us, and why people still matter when they get there. Long-form thinking, stored in the stacks below.