Anyscale, the company behind the open-source Ray distributed compute framework, has flipped its managed AI platform into public preview on Microsoft Azure. The June 2, 2026 launch makes Anyscale on Azure a first-party native integration delivered through Azure Resource Manager and running on Azure Kubernetes Service inside each customers tenancy.
Sovereign AI Inside The Customer Tenant
Anyscale on Azure is the result of co-engineering with Microsofts Azure Cloud Native team and arrives as enterprises hit the limits of relying solely on third-party model APIs. Workloads run inside the customers own Azure account, with Microsoft Entra ID, role-based access controls and audit trails carrying through to AI pipelines. Anyscale CEO Keerti Melkote described the platform as built for teams that have decided their AI is core enough to own, while Microsoft Technical Fellow Brendan Burns said the integration brings the popular open-source Ray engine directly into Azure for customers that want to keep proprietary data inside the cloud they already trust.
One Runtime For Data, Training And Inference
Rather than stitching CPU-oriented data engines together with hosted GPU APIs, Anyscale on Azure exposes a single runtime for multimodal data preparation, distributed training, fine-tuning, reinforcement learning, batch inference and agentic workloads. Anyscale says early customers are seeing up to 4x faster experimentation and up to 90% lower AI total cost of ownership compared to fragmented stacks that combine cloud-native data platforms with hosted model APIs.
Wayve, Xoople And A Sovereignty Push
Geospatial AI company Xoople is using Anyscale on Azure to process planetary-scale satellite imagery, while UK self-driving startup Wayve is running large-scale distributed ML and data pipelines on the platform across CPU and GPU fleets. Anyscale consumption draws down against existing Microsoft Azure Consumption Commitments, meaning enterprises can route AI spend through budgets already in place. The launch parallels Microsofts agentic AI push at Build 2026 and the broader AI infrastructure race.
Reporting based on coverage from Anyscale, Microsoft Tech Community, PR Newswire and Futurum Group.