HPE took to the stage in Las Vegas this week, and promptly dived deeper into AI, sewing the agentic version through its offering and explicitly targeting “sovereigns” desperate to close the AI gap.

The giant kicked off its Discover event by unwrapping a revamp of its Greenlake platform promising that agentic AI will allow IT admins to catch up on their sleep.

The firm has also integrated three key acquisitions it made in recent years – OpsRamp, Morpheus and Zerto under the CloudOps banner.

And it has expanded its “AI factory portfolio”, rounding out what CTO Fidelma Russo described as “one bold visions for the future of AI”, which it will be pitching at governments desperate to close the AI gap with the US.

HPE’s Greenlake hybrid cloud platform was launched in 2018. Which, coincidentally was around when AI erupted back into the mainstream.

HPE SVP and GM for OpsRamp Software and Cloud Platforms, Varma Kunaparaju called attention to “The way AIops evolved over the last six, seven years.” Initially, it meant ingesting large amounts of alerts and data to “really figure out a signal out of that noise.”

Then the emergence of chabots allowed teams to ask questions in natural language and “get clear relevant answers.” But, he continued, “The third part of the journey is all about agentic AI, and how to use that to transform operations, not just optimize them.”

He said the company was now building agents “grounded” in full stack telemetry courtesy of “the OpsRamp observability” across its specific domains. Model context protocol will be part of this.

“These systems are powered by additional language models that include a mixture of domain specific experts that will ultimately result in planning, orchestrating and root cause analysis for proactive mitigation”

He claimed this would deliver outcomes that matter, ie, fix faster and fail less. “It will be fully autonomous” though “in certain cases a human in the loop is necessary.”

General models need not apply

Generalist language models were not sufficient to support these applications, he argued. “Unless one understands the full stack from application all the way to the infrastructure, the ability to kind of really do the intelligence does not exist.”

Greenlake will underpin agentic tooling for its Aruba Networking Central network management system. SVP and CPO Aruba Networking David Hughes, added that the firm had been collecting telemetry “from our customers’ networks 24 by 7 for almost a decade.”

Now, he said, “We are able to have a autonomous supervisory AI module talk to many different specialized sub agents.” This could analyze the data, pinpoint root cases and figure out recommendations.

“It's like having a teammate that can work on problems while you're asleep, and when you arrive in the morning, have those proposed answers there, complete with chain of thought logic, explaining how they got to their conclusions."

In parallel, HPE is more tightly integrating its Zerto, OpRamp and Morpheus tools. HPE bought Zerto in 2021, OpsRamp in 2023, and Morpheus in 2024. It has been progressively integrating them with other parts of its broader technology offer.

Now it has rebranded the three as a “suite”, spanning resilience and data protection,, automation and orchestration, and governance, though it will continue to sell them separately.

Meanwhile, it has announced an integration covering Morpheus  VM Essentials with Veeam, and an expanded partnership with Commvault, that will see the company offer Zerto.

And the firm played up its AI Factory strategy, touting “modular” solutions, powered by NVIDIA Blackwell, with NVIDIA validated designs, and featuring its Alletra Storage MP platform. The latter is now also touted as offering “AI-ready data through full integration of Model Context Protocol”.

The firm singled out “sovereign solutions” as a target market. Non-US governments are pledging large amounts of cash for AI infrastructure as they try and avoid being squeezed by China and the US.

Trish Damkroger, SVP and GM for HPC and AI Infrastructure, described this as “an emerging market where national governments and public sector entities, [are] looking to invest in their own infrastructure, data and workforce.”

She said “It's the science and research that is supporting the national security mission. So it's critical that this infrastructure and the solutions comply with all the government and governance policies.”

She said HPE’s stack would deliver “The governance and control required to protect the propriety data and IP and be compliant so at the infrastructure layer.”

It’s something HPE needs to get right. Its most recent results were dented by its server business, with CEO Antonio Neri saying, “We delivered a solid quarter but we could have executed better.” It had been forced to re-evaluate its traditional server inventory, and was hit by “higher than normal AI inventory” as demand shifted to newer GPUs and other components.

Join peers following The Stack on LinkedIn

The link has been copied!