Hewlett Packard Enterprise (HPE) says advances to its InfoSight predictive-analytics platform is the “next step toward an autonomous data center.”

HPE acquired the InfoSight technology from Nimble Storage, which it bought for $1 billion earlier this year. It uses artificial intelligence (AI) to analyze millions of sensors across the globally connected installed base — HPE puts the number at more than 10,000 Nimble Storage customers — and uses this data to help predict and prevent problems.

The company is adding an AI recommendation engine to the platform, which will be available to all Nimble Storage customers in January 2018. This new feature builds on InfoSight’s predictive capabilities and will preemptively advise IT departments how to avoid issues, improve performance, and optimize resources. These recommendations are based on data science from machine learning and telemetry collected from customers.

In addition to adding new machine learning capabilities, HPE is also expanding InfoSight across its 3PAR storage portfolio. HPE InfoSight for 3PAR will also be available in January 2018 to all 3PAR customers. This first release will include cross-stack analytics and global visibility to customers across all of their 3PAR arrays.

Both Nimble and 3PAR are HPE primary storage products that include all-flash arrays.

“This is putting us on the path toward an autonomous data center,” said HPE’s Vish Mulchand, senior director, product management and marketing, storage division. “The recommendation engine is centered around giving the customer more information, telling the customer what to do to make their environment better. For example, we can recognize something that’s not configured correctly and isn’t yet causing a problem — but might rear its head down the road. This will avoid future problems.”

But to get to a fully autonomous data center, the technology needs to move from a recommendation engine to a command that feeds the operation. The first step, Mulchand said, would be for it to suggest actions based on the recommendations.

“The same recommendation engine could say ‘we recommend you do A, B, and C for these reasons; would you like us to do it for you?’ And then the recommendation would drive the actual operation that needs to be performed.”

Additionally, the technology needs to cover the entire data center infrastructure.

HPE plans to extend the machine learning capabilities across its entire data center portfolio, Mulchand said.

“Today InfoSight stems from the storage. It does look out across the data center infrastructure, but from a storage perspective HPE is in the perfect position to bring these services to servers, networking equipment, across the entire portfolio of hybrid IT solutions we sell.”

So what data center product is up next? While the company hasn’t announced anything, Mulchand said, “We have Simplivity hyperconverged, that often comes up in conversation." HPE this quarter launched its first hyperconverged infrastructure (HCI) system integrated with Simplivity software, that it acquired for $650 million in January. “The servers would be a no brainer for HPE. If you had to guess — and this is not a roadmap indication — those would be very important candidates."