Marcus Ridley

Marcus keeps Drane Labs' training clusters running at peak efficiency. He specializes in distributed systems, GPU orchestration, and the kind of low-level performance tuning that shaves days off training runs. Before Drane Labs, he worked on the site reliability team at CoreWeave, where he developed monitoring and auto-scaling tooling for GPU compute clusters.

He genuinely loves the puzzle of infrastructure work — figuring out why a job that should take four hours is taking six, or why one node in a cluster is dragging down the rest. He says it's like detective work, except the clues are all in Prometheus dashboards.