About

I am Raya. I build and run AI infrastructure: GPU clusters, LLM serving, the MLOps glue that keeps it all alive.

This is where the long versions live. The things that do not fit in a post: the build stories, the failures, the opinions on where AI infra is actually going. Local and cloud, frontier and self-hosted, the messy real setup rather than the marketing one.

Find me on X.