Data infra / DBs / DistSys -> ML infra + inference systems engineer in SF. Fan of cortados, Man United, Steph Curry, dogs, the Wiggle, and jazz.
Work
Engineering / LLMs at Anyscale working on batch inference, large scale serving, and Ray.
Previously helped launch Workers AI at Cloudflare, and worked on
exascale compute and search / ML data infra at Apple under Siri Search.
Writes
Occasional writing on AI, infrastructure, and policy.
Acemoglu & Johnson, Power and ProgressTechnology doesn't automatically produce shared prosperity. It's a political question.
Rodrik, “Premature Deindustrialization”Named the phenomenon: economies skipping manufacturing without building institutional capacity.
Baldwin, The Great ConvergenceHow ICT built the offshoring model. Explains why AI is a structural reversal, not a cycle.
Obermeyer et al., “Dissecting Racial Bias in an Algorithm”The canonical example of proxy variables embedding bias. Changed how I think about training data.
Sunstein, SludgeAdministrative burden as inequity. Complexity itself is a policy instrument.
Selbst et al., “Fairness and Abstraction in Sociotechnical Systems”Why fairness can't be solved as a purely technical problem.