Architect and build SkyPilot's commercial managed control plane, focusing on the API server, multi-tenancy, and enterprise-grade platform services. Ensure the reliability, scalability, and security of the hosted infrastructure to support large enterprise GPU fleets.
About SkyPilot
SkyPilot accelerates the world's most ambitious AI teams. Every hour they spend fighting infrastructure is an hour the frontier doesn't move — so SkyPilot turns fragmented compute across clusters into one optimized, highly available and easy-to-use pool: a single "AI supercomputer."
SkyPilot (10k+ GitHub stars, 14M+ downloads) is deployed at 100s of companies — from Fortune 500s to top AI-natives like Abridge, Applied Compute, Mistral, Unconventional AI, H Company, and Nubank — with usage growing exponentially. Born in the UC Berkeley lab behind Spark and Databricks, our growing team includes top-tier talent from Databricks, Google, Berkeley, MIT, CMU, and Cornell.
The role
SkyPilot Platform is a managed control plane on top of the open-source core — a hosted API server that gives infrastructure teams unified scheduling, governance, and security across all their clusters and clouds, while compute stays in their own environment (BYOC). We're looking for an engineer to own this control plane end to end — the API server, multi-tenancy, and the services that turn SkyPilot into a product teams run their infrastructure on. This is the layer that makes a large enterprise comfortable putting its whole GPU fleet on SkyPilot.
What you'll do
Architect SkyPilot's commercial platform from the ground up: the managed API server (hosted control plane), control-plane / data-plane separation, and the multi-tenant foundation teams run their compute on.
Scale and operate the control plane: keep the hosted API server reliable and highly available — monitoring, alerting, seamless upgrades, and the observability teams depend on.
“I was the first applicant for a remote marketing position that got listed on the company website the same day I applied. Had an interview within 48 hours!”