Kathy Gibson reports from Nutanix .Next in Washington DC – Use cases for artificial intelligence (AI) are rolling out quickly across many organisations – and Nutanix has been quick off the mark in introducing its own AI-based solutions.

Manosiz Bhattacharyya, chief technology officer of Nutanix, explains that 2024 was the year that generative AI (GenAI) took off, with enterprises harnessing the technology for numerous solutions – and 2025 is quickly becoming the year of agentic AI.

With Nutanix Enterprise AI, Nutanix keeps AI costs predictable, and provides the platform for customers to accelerate their AI implementations.

“But Nutanix Enterprise AI is not just something we are selling to customers,” Bhattacharyya says. “We are also internalising it within Nutanix ourselves.”

He explains how Nutanix has evolved over the years, from a user interface-centric company to an API-centric supplier, offering a consistent, secure experience.

“In the future, with Nutanix Enterprise AI and APIs, we can create agents for many applications. These will include the ability to monitor Nutanix for you and take actions on your behalf.

“This is what autonomous data centres are all about – and Nutanix Enterprise AI will enable it.”

The key to Nutanix’s agentic AI value is the fact that it runs everywhere, Bhattacharyya says. “We are at the edge, in the data centre, and in the cloud.

“Data is being generated everywhere: imagine we can take AI to that data instead of having to bring the data to AI?

“If we could run at the edge on bare metal, in the data centre on a hypervisor, and cloud-native in the cloud this will offer tremendous efficiency and consistency in managing data across all locations.”

This can be further improved with Nutanix Unified Storage, Bhattacharyya adds. “NUS is not an appliance – it’s a software application, and so is Nutanix Enterprise AI.

“If we can run both of them on commodity servers, which can run GPUs, we can bring AI as close to the data as possible. And the storage itself can automatically become smart.”

Ashwini Vasanath, principal product manager at Nutanix, says agentic AI can be used to monitor the Nutanix environment and suggest actions.

She demonstrated some use cases including automatic translation of system documentation and a compliance check for content – the latter performed by the storage system.

In addition, an AI agent was set up to monitor system performance: when it exceeded an acceptable range, the agent suggested an action and carried it out without any human intervention.

“2024 was about managing models with Nutanix Enterprise AI,” she says. “2025 will be about allowing us to enable agents using these models.”

In the future, Vasanath adds, enterprises will have to consider how to manage the agents themselves, ensuring they are secure and able to work seamlessly with staff and with one another.

Bhattacharyya stresses that all the agentic AI features demonstrated today at Nutanix .Next were enabled by Nutanix Enterprise AI, and customers can build these or similar solutions themselves using tools available today.

Lee Caswell, senior vice-president: product and solutions marketing at Nutanix, believes AI has the potential to change the way companies interact with their customers.

Implementing AI solutions is not without its challenges, Caswell says. “So we are seeing customers deploy AI first in their internal operations before exposing it externally.”

As an example, he suggests that organisations would look to save time and improve efficiency as the justification for deploying AI.

For instance, support staff could use GenAI to quickly access more information and a specific issue, which they could then apply their own judgement to.

“Today, there is a lot of manual reviewing of documents before an agent can present a solution to the customer,” Caswell says. “AI could be synthesising those documents for them. This way, the support team gets more complete information.”

Adding agentic AI – with guardrails – could allow the system instead of a human to refine those results agentically.

“That internal use of AI is very effective and can be financially justified.”