$ sky mission --
Our mission is to reimagine the cloud to accelerate the world’s most compute-intensive workloads
About SkyPilot
SkyPilot is the open source platform for running, managing, and scaling AI workloads on any infrastructure
SkyPilot started at the Sky Computing Lab at UC Berkeley, where our founders were studying the future of cloud infrastructure. The challenge was clear: today’s cloud is incredibly powerful, but also incredibly complex. Every team faces the same maze of decisions: which cloud? which region? which GPU? reserved or on-demand? Kubernetes or VMs? Each choice multiplies the complexity, and getting it wrong can mean wasted resources, ballooning costs, or missed deadlines.
To solve this, our team proposed a new idea: Sky Computing. Instead of forcing organizations to pick one cloud, one cluster, or one vendor, why not unify all of them into a single pool — a “sky” of compute? That vision became SkyPilot. First as a research project, then as an open source platform shaped by contributions from engineers, researchers, and AI teams around the world.
Trusted by AI teams, everywhere
SkyPilot in numbers
Built by researchers, for researchers
Backed by leading investors and operators
OUR ALL-STAR ANGELS INCLUDE
Come build the future of AI infrastructure
We’re growing our team of researchers, engineers, and community builders. Explore open roles and help us chart the future of AI infra.
Loading job opportunities...
