About Me
I'm a cloud economist and ML engineer working on the economics of compute — what it costs, in dollars and in watts, to deliver a unit of intelligence.
My foundation is FinOps across AWS, Azure and GCP: turning complex cloud and infrastructure spend into executive-ready decisions, and embedding cost accountability into daily IT operations. But the metric the industry is converging on isn't FLOPS — raw processing speed on a spec sheet — it's the cost and energy behind every answer an AI system produces. So the same discipline now runs further down the stack: inference economics at the bill, hardware choice at the box, thermal footprint at the building, and ML-driven optimization of the silicon underneath it all. Watts and dollars per token, from the cloud bill to the chip.
A physics background lets me operate credibly at both ends of that metric — the cost layer and the energy layer — and I build the ML systems myself rather than only advising on them. I trace problems back to root causes, asking why and how outcomes emerge rather than stopping at what's visible — which surfaces structural inefficiencies instead of symptoms.