Anyscale, the creator of the Ray open-source project, has announced the public preview of Anyscale on Azure, introducing a native AI compute platform designed to help enterprises build, train, and scale AI workloads directly within their own Microsoft Azure environments.
Built on Azure Kubernetes Service (AKS) and Azure Resource Manager (ARM), the platform enables organizations to manage AI workloads using the same security, governance, identity, and billing models already used across their Azure infrastructure. The launch reflects a growing enterprise shift toward sovereign AI strategies, where organizations seek greater control over AI infrastructure, proprietary data, and long-term operational costs.
As enterprises transition from AI experimentation to production deployment, many organizations are reevaluating their reliance on externally hosted AI APIs due to rising operational costs, governance concerns, and the need for greater control over proprietary data.
Anyscale on Azure addresses these challenges by enabling enterprises to run AI systems entirely within their own cloud infrastructure. The platform allows businesses to develop custom AI models, process large-scale multimodal datasets, and deploy AI applications without moving sensitive data outside their controlled Azure environments.
"AI has quickly become one of the largest and least predictable line items in the enterprise IT budget," said Keerti Melkote, CEO of Anyscale. "The companies pulling ahead are not necessarily spending less on AI. They are gaining more control over how that spend scales. Instead of only renting intelligence through APIs, they are building and operating AI systems inside their own cloud. Anyscale is for the teams that have decided their AI is core enough to own."
The solution is positioned as a unified compute foundation for the complete AI lifecycle, including data processing, model training, fine-tuning, inference, and agentic AI operations.
Delivered as an Azure Native Integration, Anyscale on Azure operates fully inside customer-owned Azure tenancies. The platform inherits Azure’s native security, identity management, and governance capabilities, allowing enterprises to apply existing Microsoft Entra ID policies, role-based access controls, audit systems, and compliance frameworks directly to AI workloads.
This architecture is particularly relevant for organizations operating in highly regulated industries such as financial services, healthcare, and government sectors where data residency, governance, and infrastructure sovereignty remain critical requirements.
"There's growing interest from enterprise customers in building AI inside their own Microsoft Azure cloud environment, on their own data, with more control over how costs scale," said Brendan Burns, Technical Fellow and CVP, Azure Cloud Native, Microsoft. "Anyscale on Azure brings the popular open-source Ray engine directly into Azure, giving customers a great option to build and operate AI systems within their existing Azure environments."
Anyscale is built on Ray, the open-source distributed compute framework designed to scale Python and AI workloads across CPU and GPU infrastructure.
The platform supports distributed multimodal data processing, reinforcement learning, inference workloads, and large-scale model training while helping organizations reduce infrastructure fragmentation and improve GPU utilization efficiency.
According to Anyscale, enterprises using the platform can replace unpredictable per-token AI API pricing models with infrastructure they directly govern and optimize internally.
The company states that customers have reported up to four times faster experimentation cycles and up to 90% lower AI total cost of ownership compared to fragmented AI stacks relying on separate cloud-native processing tools and externally hosted model APIs.
Several organizations are already leveraging the platform for large-scale AI operations.
Geospatial AI company Xoople is using Anyscale on Azure to process planetary-scale satellite imagery and accelerate AI deployment workflows.
"With Anyscale on Azure, Xoople can reliably run massive AI workloads over planetary-scale satellite imagery, transforming complex spectral data into decision-ready intelligence," said Milos Colic, VP of Engineering at Xoople.
Autonomous driving company Wayve is also expanding its AI infrastructure using Anyscale on Azure to support distributed machine learning pipelines and GPU-intensive autonomous driving model development.
The growing adoption highlights increasing enterprise demand for scalable, governed AI infrastructure capable of supporting production-grade AI systems while maintaining operational control and cost visibility.
Anyscale’s collaboration with Microsoft signals a broader industry movement toward enterprise-owned AI ecosystems built on open-source frameworks, sovereign infrastructure models, and scalable cloud-native architectures designed for long-term AI operations.
About Anyscale
Anyscale is the AI compute platform built by the creators of Ray, the most widely adopted open-source framework for scaling Python and AI workloads. Anyscale powers AI at companies including Coinbase, Bedrock Robotics, and Runway, and is used to train, fine-tune, serve, and process multimodal data for some of the largest AI systems in production.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones, and thrive on the tech pulse.
