SkyPilot, an AI compute management platform that emerged from UC Berkeley research, raised $20 million in seed funding and launched SkyPilot Platform, a commercial product that lets organizations run AI workloads across cloud providers and on-premise clusters through a single control plane.
Lux Capital led the seed round, with Amplify Partners, Coatue Management, Foundation Capital, Race Capital and The House Fund also participating. The investor group includes Ali Ghodsi, CEO of Databricks; Jeff Dean, Chief Scientist at Google; Guillermo Rauch, CEO of Vercel; Amjad Masad, CEO of Replit; Clem Delangue, CEO of Hugging Face; and Tristan Handy, CEO of dbt Labs.
Zongheng Yang, Zhanghao Wu and Romil Bhardwaj, all Berkeley researchers, founded SkyPilot alongside Ion Stoica, a Databricks co-founder and Berkeley professor, and Scott Shenker, a networking and cloud computing researcher also at Berkeley.
Hundreds of organizations run on the platform, including Nubank, Abridge, Applied Compute, H Company, Archer and Hippocratic AI, spanning AI research, healthcare and financial services. Before the commercial launch, the project had reached 14 million downloads and attracted more than 280 contributors.
"Every organization is building custom intelligence around its own data and domains," said Zongheng Yang, CEO and co-founder of SkyPilot. "The challenge is that the AI compute needed to build it is fragmented across clouds. SkyPilot gives frontier AI teams a single platform to manage that infrastructure so they can build custom intelligence faster."
SkyPilot Platform adds fleet-wide GPU health monitoring, automated remediation and SOC 2 compliance on top of the open-source foundation. The commercial version handles workload orchestration across training, inference, reinforcement learning, evaluations and multi-cluster production serving, along with high availability, team and quota management, and single sign-on.
Private preview customers managed more than 10,000 GPUs and supported more than 200 researchers per organization. Some recorded performance improvements of up to 20x over the open-source version, which the company attributes to platform-level optimizations.
"Every organization building custom intelligence is running into the same constraint. The compute they need exists, but it's fragmented across clouds, clusters, and hardware generations," said Brandon Reeves, Partner at Lux Capital. "SkyPilot brings that infrastructure together into a single operating environment, helping customers achieve double-digit improvements in GPU utilization while giving frontier AI teams the foundation to build and scale custom intelligence."
The $20 million will fund product development and engineering, with continued investment in the open-source ecosystem. Access is now open for select new customers at skypilot.ai.
