Hi, I’m Preetham.

I build AI systems that act on your behalf without taking you out of the loop. Most of my work sits at one of two edges: agents with real autonomy that still ask before they act, and on-device models that keep sensitive data on the device.

Across 15 projects — browser agents, multi-agent orchestrators, on-device health and accessibility tools, and creative-AI pipelines — the through-line is human-supervised autonomy: approval gates, audit trails, and privacy as architecture rather than promise.

When I’m not building, I’m reading — this site’s research feed is my running log of the frontier, and the writing is where I think it through.