About

I'm Caaren Amirian — an AI engineer working on one problem: making autonomous systems trustworthy. For me that's concrete, not a slogan. It means robots that behave predictably in simulation before they touch hardware, and agents whose every decision can be inspected, reproduced, and rolled back. Verifiable, replayable, fail-closed.

My work spans robotics simulation (NVIDIA Isaac Sim, ROS 2, edge inference on Jetson) and agentic verification — the workflows that check whether an AI system actually did what it claimed, with evidence. The thread through all of it is accountability: I'd rather ship a system that can prove it's right than one that's merely confident.

I build in public. If you care about AI that's accountable — or you're hiring for it — reach out.