.NEXT 2026 - why Nutanix CEO Rajiv Ramaswami is betting on agentic AI being a hybrid enterprise application

Nutanix kicked off its annual .NEXT 2026 user conference  in Chicago with a series of announcements extending its platform for the agentic AI era. As per the corporate blah blah, these included:

• The Nutanix Agentic AI solution, a full-stack platform announced during NVIDIA GTC 2026 and currently in early access, is designed to help enterprises build and operate AI applications on NCP. The full solution will be available in the second half of 2026 and include a secure, high-performance virtualization foundation for AI infrastructure, and integrate compute, storage, networking, and Kubernetes services to simplify deployment and operations. Together, these capabilities will enable enterprises to run modern and AI workloads efficiently across hybrid and multi-cloud environments.

• NKP Metal, announced today, is in early access and will be generally available in the second half of 2026.  It extends the Nutanix Kubernetes Platform (NKP) solution to support Kubernetes deployments directly on bare-metal infrastructure, delivering the performance for edge environments and AI training workloads that rely on dense GPU infrastructure.

• Nutanix Unified Storage (NUS) 5.3 is generally available now and ideally suited to drive the transformation of object storage into a performance storage tier required for AI Factories. The release expands Smart Tiering to enable seamless data movement to Google Cloud and OVHCloud S3, while adding multi-tenant object scaling and quotas to support massive AI data lakes. NUS will also introduce, later in 2026, Remote Direct Memory Access (RDMA) acceleration for S3-compatible object storage to dramatically increase throughput for large AI training datasets and data-intensive pipelines.

• The updated Nutanix Data Lens 2.0 solution is generally available now and can run fully on-premises, including in air-gapped environments. The release brings ransomware analytics, data audit and governance, and visibility across distributed storage footprints to sovereign and dark-site deployments that cannot rely on SaaS-based data security.

• Nutanix and MongoDB announced a certified integration, generally available now, between Nutanix Database Service and MongoDB Ops Manager that is built on MongoDB’s third-party backup integration model. Nutanix and MongoDB are collaborating to simplify enterprise database operations with automated provisioning and lifecycle management across infrastructure and database environments.

CEO perspective

At an analyst side event at the gig, CEO Rajiv Ramaswami drilled down into the company’s strategy in helping users navigate the shift towards enterprise-scale agentic adoption. His pitch:

Nutanix truly delivers a unified modern platform, powering the apps of tomorrow, the AI apps of tomorrow, but also the mission-critical apps of today, enabling our customers to use us across a wide variety of fronts, running their existing business, modernizing everything and innovating in the AI future.

There’s a basic customer proposition in play here, he argues: 

As we talk to customers today, there's a lot that's on their minds. On the one side, every CIO that I talk to, every customer that I talk to, is thinking about how can they operationalize AI in their enterprise while dealing with the complexity that it brings to the table. There's AI in the public cloud, AI on-prem, AI everywhere, and they struggle to figure out how to deal with it and how to operationalize it in their companies and get tangible ROIs on it. The AI factories are here, but then again, putting it all together to make this thing work for them is no easy task. At the same time, we've got the geo-political situation that we're all sitting in. What that means for us as a company is that there is a lot more focus on sovereignty. 

So far, so pretty familiar pitch. What does Nutanix bring to the party that others don’t? Ramaswami argues:

The value proposition that [customers] see from Nutanix - we deliver them the simplicity of experience while giving them great total cost of ownership on the other side. We give them the flexibility to use us for many different use cases and give them at the same time, the control, the security they need and the performance they need for their business critical applications, all the while making sure we support them fantastically, in a fantastic way.

Customers buy into Nutanix for the long haul, he adds:

They start with us, and they continue to grow with us. What Nutanix does uniquely, I would say more so than anybody else in the industry, is that we deliver that single unified platform for today and for tomorrow...If customers want to modernize their infrastructure, we have a solution for them. They want to re-use their existing hardware while modernizing, we have a solution for them. They want to run in the public cloud, we have a solution for them. They want to modernize their applications, go to a cloud native framework, we have a solution. And now with our agentic AI platform, we're enabling them to run their agentic AI applications.

Agentic hybrids

Agentic AI is going to be a “true hybrid application” for most enterprises, he predicts:

There's going to be applications that run in the public cloud, there’s going to be applications that run in the private cloud and the edges, and there's going to be many applications that run in these so-called neo-clouds, which are a whole host of new service providers that provide AI services. We are focusing on capturing the opportunity across all of these.

The reason that AI will be hybrid is, again, if you look at the private cloud and the edge, you've got sovereignty being a big push. You've got regulation being another big push. You've got the proximity to data being another reason. You've got the need to do real-time inferencing for a lot of these new use cases that are being -- coming up at the edge and in these manufacturing sites and other places. And then there's customers who are going to be consuming this, of course, in the public cloud and in neo-clouds as well. They can get there for a good subset of applications, that will also be an option. We expect that the slew of AI applications will continue to be hybrid, just like today's world is hybrid.

The rise of AI factories helps to some extent here, he argues, but end user needs are moving on:

When we started with AI factories, they were specialized elements. They were serving the needs of a sub-set of users, kind of a small portion of the enterprise. But this is exploding. It's exploding because of scale at which people are now building and deploying these applications is just tremendous. You're servicing more and more business users. You're servicing more and more developers, AI engineers and the number of agents is exploding.

This leads to a problem:

You've now got these AI factories that deliver critical infrastructure for all these needs. They need to be operated by the infrastructure admins and the platform engineers..It’s a constrained resource. We need to optimize it. We need to provide security, governance, all of these things.

All of these things Nutanix already does for the compute-centric world, and now it aims to do the same for the AI inferencing and agentic world, boasts Ramaswami:

What we do in very simple terms is to make this a cloud operating model for these AI factories. We provide a turnkey experience so that customers don't have to do the work of integrating everything, and they can start being consumers of infrastructure for their AI use cases rather than trying to go put it all together and run it all themselves.

The Nutanix stack underpinning this has a number of core elements - a set of AI services with an underlying Kubernetes platform to run them; a data foundation to stream the data that AI applications need with low latency and high performance; and the ability to manage all the shared infrastructure across multiple tenants and multiple users.

Ramaswami argues:

Essentially, what we deliver with this cloud operating model is a turnkey platform that allows companies to go build and run their AI apps with all the stuff that they need to do so, the security, the control, the governance, being able to manage all of this. It’s all the best performance and drive to the lowest cost per token, which is a unit of intelligence.

He cited an unnamed EMEA-based sovereign digital services provider as an end user exemplar of how all this works in practice:

Sovereignty is very important for them. They are a digital services provider to many people, and they have been a customer for several years now. They started out with a standard use case with us. Modernizing their HCI (Hyper-Converged Infrastructure), they ran all their databases on our platform. And then over time, they consolidated the vast majority of their enterprise applications onto our platform. That was the second stage. And then in the third stage, they're now deploying Nutanix to create a shared AI infrastructure for their multiple tenants. So that again, they can provide a shared service to all their tenants, maximize the utilization of their shared infrastructure and deliver this in a secure way using our agentic AI stack.

My take

Truly, we are the platform of the future.

Well, time will tell on that bold prediction, but it was a compelling sales pitch and few would question the long-term loyalty displayed by the Nutanix customer faithful.

Mark Chillingworth is on the ground at .NEXT. More to come from him. 

Onwards!