Nutanix CEO on treating AI as core infrastructure

Rajiv Ramaswami urges IT leaders to prioritise AI governance, cost control and platform adaptability as agentic AI ramps up

Image: Rajiv Ramaswami, CEO, Nutanix on stage at the Nutanix .NEXT 2026 conference

As agentic AI moves from pilot into production, CIOs face mounting pressure to govern cost, data, and risk - forcing a rethink of not just workloads but platforms too.

At the Nutanix .NEXT 2026 conference on Monday in Chicago, the message to enterprise IT leaders was clear that agentic AI is no longer a future concept. It is now an operational reality.

Yet as AI moves from pilot projects into production, it exposes hard questions surrounding governance, cost control, data sovereignty and infrastructure flexibility.

In his opening keynote, Nutanix CEO Rajiv Ramaswami and several of Nutanix’s enterprise customers spoke on why AI adoption is forcing IT leaders to rethink their operational platforms, not just their workloads.

Here are some key takeaways from Nutanix’s CEO and its customers:

Agentic AI is moving into governed production

Ramaswami described a shift away from simple AI prompting toward autonomous agents that operate continuously across the enterprise.

“We are rapidly moving from an era of prompting to the era of delegating and empowering autonomy with agents,” Ramaswami said. “This is now about your competitive edge.”

He emphasised that organisations are now dealing with more users, more agents and more data - often with limited infrastructure resources. As a result, AI must be treated like core infrastructure, requiring standardised platforms and operational controls rather than fragmented tools.

Governance and data sovereignty are mandatory AI requirements

The keynote and customer stories made it clear that regulated industries can’t let AI run amok. Clear rules on where data lives, who can access what, and how AI is allowed to operate are mandatory as AI spreads.

That concern is already shaping real-world decisions. Dan Regalado, CIO of the high-end leisure chain Wynn North America, who joined Ramaswami on stage at the NEXT event, emphasised that regulatory constraints and data residency both define how AI can be deployed.

“Our gaming data cannot leave the state,” Regalado said. “Data security and data residency are non-negotiable for every one of our resorts.”

Even organisations experimenting with AI in the public cloud are reassessing where production workloads belong.

“We’ve done it in the cloud,” Regalado added, “but we’re actively researching how we might do it better on‑prem or in a hybrid model, because governance and control matter.”

Whilst Wynn is in the heavily regulated casino industry, IT leaders in other verticals must assume regulators, auditors and boards will demand clear answers on their AI strategy.

AI cost predictability has become an IT operations issue

As AI use scales, tracking usage and managing costs are no longer just financial concerns.

Nutanix highlighted the growing importance of usage metering and “cost per token” visibility to prevent AI initiatives from becoming budget liabilities.

“As you use more and more models, you run into challenges around tracking usage and managing cost,” Ramaswami said.

For Wynn, cost efficiency is already influencing platform decisions.

“Cost efficiency is one of the major factors we’re evaluating,” Regalado said, “as we decide whether to stay purely in the cloud or invest on‑prem.”

AI without metering also becomes a budget liability. CIOs need infrastructure that surfaces usage clearly and supports predictable budgeting.

Platform simplicity matters even for smaller IT teams

Customer examples showed that many IT teams, especially in the midmarket, often lack the staff to manage a growing mix of AI tools. The customers conveyed a very specific message: Reducing operational complexity through unified platforms is critical to sustaining innovation without expanding head count.

For many SME’s, AI ambition can clash with limited staffing, said Josh Hostetler, lead platform engineer at online tyre and wheel specialist, Tire Rack.

“I’m on a platform engineering team of three people,” Hostetler said. “Our goal was to reduce administrative burden without adding another tech stack or more engineers.”

Tire Rack modernised incrementally, starting with foundational workloads and evolving toward containers, without overwhelming the team.

“We started simply,” Hostetler said. “Stability mattered as much as scale.”

AI platforms that increase operational complexity can halt adoption in thIs type of organisation.

AI is forcing platform adaptability

From supply chain constraints to cloud cost concerns, IT leaders are evaluating where workloads should run. Hybrid and on‑premises options are gaining renewed attention as AI demands more platform adaptability.

Stephen Hall, vice president of infrastructure and operations at BlueCross BlueShield of Tennessee, underscored the importance of adaptability as AI reshapes infrastructure planning.

“Infrastructure leaders need adaptability,” Hall said, “because the industry will keep evolving.”

This article first appeared in our US sister title MES Computing