Mirantis and VAST Data are combining k0rdent AI with the VAST AI Operating System to provide AI cloud operators with an automated platform for secure multitenant infrastructure. The integration coordinates compute capacity, storage resources, and data services through a common orchestration model. It is intended for neoclouds, service providers, and GPU as a service operators seeking faster customer onboarding, lower administrative effort, and scalable AI workload support for demanding customers.
Table of Contents: What awaits you in this article
Mirantis VAST platform integrates GPU compute and AI data
The integration combines Mirantis k0rdent AI infrastructure orchestration with data services from the VAST AI Operating System. It establishes an operational framework that manages GPU computing resources alongside the data layer required by artificial intelligence workloads. The approach targets cloud service providers, neocloud operators, telecommunications companies, and GPU as a service providers. These organizations can use the combined platform to build, operate, and scale AI infrastructure for customers efficiently and securely.
k0rdent AI Automates Tenant Storage Across Compute Lifecycle Operations
k0rdent AI automates the delivery of tenant-isolated VAST data services alongside compute capacity, allowing storage to be created, expanded, and retired as part of a single infrastructure workflow. GPU clusters, Kubernetes components, and data services follow the same lifecycle rules. When an operator deploys or enlarges a customer environment, the required storage configuration is handled automatically, eliminating separate storage administration and reducing operational coordination across the underlying platform and complexity.
Integration Automates Isolated AI Cloud Environments Without Storage Adjustments
Deploying an AI cloud environment previously required operators to coordinate bare metal servers, Kubernetes clusters, network configuration, storage resources, and tenant specific access policies. Determining the correct network and storage parameters for each deployment required work in separate systems. Isolating individual tenant environments added complexity and increased the risk of configuration errors. The integrated platform automates creation of dedicated, appropriately sized compute and data environments without manual storage configuration adjustments.
GPU Clusters and Storage Deploy Together for Immediate Access
The integrated platform provisions GPU compute and storage simultaneously, treating both resources as components of a single deployment process. When an operator creates a GPU cluster, the associated storage capacity is automatically created, configured, and connected. Training, inference, and agentic AI workloads can therefore access required datasets immediately. This removes delays caused by separate storage requests, prevents idle GPU capacity, and provides a consistent, secure data path throughout its lifecycle.
Mirantis and VAST Standardize AI Cloud Tenant Onboarding Workflows
Mirantis and VAST present their integrated platform as a repeatable blueprint for deploying AI cloud environments. Operators can apply standardized infrastructure patterns instead of designing, integrating, and maintaining a separate configuration for every customer. According to Mirantis, this approach can reduce tenant onboarding from weeks to minutes. It also lowers operational workload by eliminating manual coordination across hardware, Kubernetes clusters, storage services, and the dependencies required for each deployment instance.
The integration of Mirantis k0rdent AI and the VAST AI Operating System addresses a fundamental operational challenge in modern AI clouds: GPUs achieve high utilization only when compute, networking, and data services are provisioned and managed together. Automated tenant isolation, a shared lifecycle for storage and compute, and standardized deployment patterns reduce infrastructure complexity for operators. This approach can accelerate environment delivery, minimize manual coordination, and provide a scalable foundation.
