VAST Data and Microsoft are jointly bringing the VAST AI Operating System to Azure. The collaboration aims to help organizations unify their data and AI pipelines. Azure users will gain access to VAST’s suite of data services, including unified storage, data cataloging, and database capabilities.
VAST Data announced during Microsoft Ignite that the AI Operating System will be available to Azure users. The platform is designed to enable the next wave of agentic AI. Organizations will soon be able to seamlessly manage data across on-premises, hybrid, and multi-cloud environments.
The VAST AI OS runs on Azure infrastructure. Customers can implement the platform using the same tools, governance, security, and billing frameworks they already know. “This collaboration with Microsoft reflects our shared vision for the future of AI infrastructure,” said Jeff Denworth, co-founder of VAST Data.
Azure users will have access to a range of capabilities. VAST InsightEngine and AgentEngine run intelligent, data-driven workflows directly where the data resides. InsightEngine delivers stateless, high-performance compute and database services that accelerate vector search, RAG pipelines, and data preparation.
Performance for model builders
For training and running models, VAST AI OS leverages Azure GPU and CPU clusters. This is done through high-throughput data services, intelligent caching, and metadata-optimized I/O. The platform utilizes new Azure Infrastructure solutions, including the Laos VM Series with Azure Boost Accelerated Networking.
An exabyte-scale DataSpace creates a unified global namespace. This eliminates data silos and simplifies data mobility. Customers can instantly scale from on-premises to Azure for GPU-accelerated workloads, without migration or reconfiguration.
VAST’s DataStore supports file (NFS, SMB), object (S3), and block protocols. The VAST DataBase combines transactional performance with the query speed of a warehouse and the cost efficiency of a data lake. This allows diverse workloads to run on a single platform.
VAST’s Disaggregated, Shared-Everything (DASE) architecture enables independent scaling of compute and storage within Azure. Built-in Similarity Reduction minimizes storage footprint and reduces costs for large-scale AI infrastructure.