Dell Technologies is looking to bring closed-loop artificial intelligence operations (AIOps) to its portfolio of hyperconverged infrastructure (HCI), compute, storage, and network products.

The company's effort includes the expansion of its CloudIQ AIOps platform and identification of five levels of machine learning-based automation ranging from AI insights to full automated remediation.

“CloudIQ, which has existed for many years, got its genesis on the storage side of the business as a way for us to provide a consistent way to look at storage health,” Drew Schulke, VP of product management at Dell Technologies, told SDxCentral. Today, “we've moved beyond storage, and we now have the full breadth of our ISG [Infrastructure Solutions Group] portfolio covered.”

CloudIQ combines monitoring, machine learning, and predictive analytics to turn health notifications, logs, and real-time reports into useful insights on the inner workings of the data center. Use cases include health monitoring, capacity planning, and risk mitigation.

According to Dell, IT teams now assess the health of their data center from a single unified dashboard, automate root cause analysis, and eventually automate remediation. “We are turning it into our autonomous operations engine,” Schulke said.

CloudIQ will eventually provide customers with five degrees, or levels of AI automation. However, for now, most customers are somewhere between level 0 (no automation) and level three (AI-generated recommendations).

“Nobody's at level five. It's there to be the North Star as we think about this framework,” Schulke said, adding that the levels allow customers to ease their way into AI automation without having to fully commit.

At each level, CloudIQ provides better recommendations and greater degrees of automation.

For example, at level one, CloudIQ might flag a storage pool that is nearing capacity for review. At this level, the platform is limited to identifying issues that might otherwise go unnoticed, Schulke explained.

At levels two and three, the platform starts making recommendations based on policy and contextual information. In the case of the storage pool, CloudIQ might recommend purchasing additional storage, or if the demand is seasonal, allocating capacity in another, underutilized storage pool.

It’s only at levels four and five that the AIOps platform starts to close the loop, Schulke said. At level four, CloudIQ will offer a recommendation alongside a one-click resolution. Meanwhile, at level five, the platform automatically resolves issues as they are detected, eliminating the human element entirely.

At levels four and five, “we're no longer having to screen and go through and look at these things and bless them,” Schulke said.

However, it isn't as simple as allowing the CloudIQ to take over, he explained. Dell isn't trying to replace data center administrator's existing tools. Instead, CloudIQ integrates with third-party applications like ServiceNow, Slack, Microsoft Teams, Ansible, and vRealize, allowing IT teams to automate data center orchestration using tools they’re already familiar with.

CloudIQ’s capabilities are still a work in progress, with levels four and five at least 18 months to 24 months out, Schulke said. He explained that not only do the algorithms need to be vetted, but even the most adventurous customers will also need time to get comfortable with AI-driven automation.

CloudIQ is available now across Dell’s complete server lineup, with additional functionality planned next month for the company’s Ethernet portfolio.