AI infrastructure isn’t just about GPUs and schedulers—it’s about the ecosystem around them. In this clip, Mirantis CTO Shaun O’Meara and VP Randy Bias explain how the k0rdent AI catalog is designed to help enterprises simplify, accelerate, and secure their AI development.

MLOps, But Modular

“Most AI workloads today are running on Kubernetes,” says O’Meara. “But what makes Kubernetes useful is everything you layer on top.”

That’s where k0rdent’s curated catalog comes in—providing observability tools, CI/CD pipelines, and security integrations out of the box. Instead of forcing users into a rigid stack, the catalog lets teams compose their own toolchains while maintaining guardrails for governance and support.

Flexible by Design

Bias reinforces the value of this flexibility: “The kind of stack you need for a visual model is very different from an LLM.” Enterprises need the ability to pick and assemble components based on the domain-specific use case—whether it’s radiology, generative AI, or autonomous systems.

k0rdent allows that flexibility without sacrificing control, offering a balance between customization and repeatability that accelerates adoption while minimizing operational drag.