Production AI Demands Infrastructure Overhaul, Nutanix Execs Warn
Breaking News — Enterprises racing to deploy artificial intelligence at scale are hitting a critical infrastructure wall, according to top executives at hybrid cloud leader Nutanix. The gap between AI experimentation and real-world production is widening as agentic systems introduce unpredictable, multi-step workflows that overload traditional data centers.
“It’s one thing to do an experiment, to do a prototype. It’s a different thing to take that prototype and deploy it for 10,000 employees,” said Thomas Cornely, EVP of product management at Nutanix. “We went from people focusing on training models to chatbots to now doing agents, where the demand and pressures on AI infrastructure are growing exponentially.”
Background
For years, organizations have run AI pilots and proofs of concept in isolated cloud environments. But as they push into production—across real workloads, real users, and regulated sectors like banking, healthcare, and government—the underlying infrastructure must evolve. Nutanix president Tarkan Maner noted that AI is reshaping not just technology but entire industries.
“AI in general is shifting everything we do, not only in technology, but across all vertical industries,” Maner said. “As a complete platform company, we welcome this change. It’s creating more opportunities for us to serve our customers in better ways.”
Agentic AI Adds New Complexity
The rise of so-called agentic AI—systems that autonomously execute multi-step tasks across applications and data sources—is driving the urgency. Enterprises now face simultaneous agents running real-time, unpredictable workloads that must coordinate across teams and infrastructure.
“OpenClaw is making it very easy now for anybody to build agents and run with agents,” Cornely explained. “You want those agents to be running on premises with your data. You need to have the right constructs around it to protect the enterprise from what an agent could do.”
These autonomous agents demand new security and governance frameworks, especially when handling sensitive corporate data. The challenge extends beyond how agents operate to how they interact with existing systems and human teams.
AI Augments, Not Replaces, Human Work
Maner emphasized that agentic AI is fundamentally an amplifier of human capability. The goal is not to eliminate workers but to find the right balance between human decision-making, AI automation, and agent workflows.
“We believe that there’s going to be love, peace, and harmony between AI, agentic tools, and robotics systems, and human capital,” Maner said. “That harmony can be optimized for better outcomes for businesses, enterprises, governments, and public sector organizations, if the right vendors provide the right tooling and the right services.”
What This Means
The shift from pilots to production forces enterprises to rethink everything from storage and compute to security and governance. IT leaders must invest in hybrid infrastructure that can handle both predictable and bursty AI workloads while keeping data on premises for compliance and low latency.
Vendors like Nutanix are positioning their platforms as the foundation for this new era. Companies that fail to adapt risk falling behind in the race to operationalize AI—and may see their experiments never reach the employees who need them most.
This is a developing story. Check back for updates on enterprise AI infrastructure trends.