Guest: Randy Bias (LinkedIn)
Company: Mirantis
Show Name: An Eye on AI
Topic: AI Infrastructure

When it comes to enterprise AI, the biggest risk isn’t failure—it’s lock-in. As standards like the Model Context Protocol (MCP) evolve at a blistering pace, organizations risk committing to technologies that may not survive the next cycle. Randy Bias, VP of Strategy & Technology at Mirantis, says that’s why flexibility is now a core design principle of AI infrastructure.

Built on Open Source, Designed for Change

Mirantis’ MCP AdaptiveOps is an open-source-based toolkit designed for one thing: adaptability. “It’s a set of tools—some we provide, some we contribute to, some from the broader community,” Bias explains. “Then we work with each customer to tailor that toolkit so it meets their needs today, and continues to evolve over time.”

That services-first approach means customers aren’t locked into proprietary systems or closed products. Whether they later decide to off-ramp to another provider or migrate to a new open-source project, Mirantis ensures a smooth transition. “If they decide to move on, they can,” Bias says. “And if we eventually release an enterprise product built from this foundation, they’ll have the option to adopt that too—on their own terms.”

Flexibility Over Finality

Bias emphasizes that no one really knows what the MCP ecosystem will look like in a few years. “Maybe it becomes huge like Kubernetes,” he says. “Or maybe it evolves differently. Right now, everyone wants maximum flexibility—and we’re committed to delivering that.”

By embracing uncertainty instead of resisting it, Mirantis is positioning AdaptiveOps as a living framework. Its modular architecture lets customers swap, upgrade, or remove components as the AI stack evolves, keeping them aligned with open standards while avoiding costly re-platforming.

The Mirantis Mindset

This open, iterative approach reflects the company’s long history of guiding enterprises through technological transitions—from OpenStack to Kubernetes to AI-native infrastructure. “We come in knowing things are going to change,” Bias says. “By designing for adaptability from the start, we make sure both we and our customers can evolve with the ecosystem.”

For enterprises navigating early-stage AI adoption, that’s more than a comfort—it’s a competitive edge.