RESOLVE ‘22 keynote review: forecasting the future of AIOps
Get info, insights, and more from our RESOLVE 2022 keynote speech.
“I kind of view AIOps as more of a practice. It is going to transform how some of these IT Operations are running today.” – Carlos Casanova, Principal Analyst, Forrester
In his opening remarks at our RESOLVE ’22 keynote, James Maguire, one of the esteemed panelists at the event, dropped an intriguing data point: “Every credit card transaction has 500 data points attached to it,” the eWeek editor-in-chief said.
What’s that got to do with the burgeoning world of AIOps? Quite a bit, as it turns out, because of the sheer complexity of the systems businesses have developed as well as the data structures they rely on to carry out even basic transactions.
“We created these enterprise IT systems that are so complex that we ourselves can no longer manage them,” Maguire said. “So we created a second complex system called AIOps to manage the first complex system.”
Maguire’s comments nevertheless speak volumes about the modern state of technology and the need for more advanced monitoring tools. Read on for more coverage of the keynote, and don’t forget to check out the entire video when you’re done with our analysis.
Exploring the need for holistic response in the new world of IT Ops
Carlos Casanova, another participant in our panel and a Principal Analyst at Forrester, added even more perspective to the matter. In the modern era, he argued, a technology outage is a lot more than “just” a technology outage. Even a small reportable event could lead to a PR nightmare or similar unsavory outcome, so businesses must take a holistic approach to matters such as error reporting.
Because of this problem, moderator and BigPanda CPO Fred Koopmans noted that “the stakes have never been higher” for businesses when systems fail. Technology failure can cause more adverse business outcomes than ever before. And that, Advent Partner Eric Noeth said, creates a backdrop with a somewhat mystifying picture of what to do next.
“That’s a perfect storm for AIOps,” Noeth said. “The problems aren’t getting easier. They’re getting harder. And it’s certainly on AIOps vendors to keep up with that complexity and rise to the challenge.”
Noeth said AIOps has quickly evolved from “a really interesting project to a necessary capability.”
“Complexity, velocity, innovation and fragmentation”: fueling the future of AIOps
While complexity, velocity, innovation and fragmentation are ultimately good things—they’re what fuel or come from the technology we rely on, after all—they can feel like a major area of added work for IT Ops teams that must navigate the ever-changing needs behind it all, Koopmans says.
According to Insight Partners Managing Director Michael Yamnitsky, those four components will also combine to create an AIOps future with even more centralization and visibility. When asked about trends he saw for AIOps in the near future, the vision he lays out is clear: a platform that gives organizations a fuller view of business outcomes and IT needs, all under a single pane of glass.
Yamnitsky also sees what he calls a shift to “domain-specific insights” soon—tools that give insights to DevOps teams, engineering teams, and other groups responsible for fixing issues. This version of the “AIOps brains under very specific domains,” he continued, would soon allow organizations to react more quickly and naturally to issues that arise—with less need for middle-manning and the information silos that come with it.
In general terms, keynote speakers said, the basic act at play here comes down to eliminating noise—making sure the right alerts go to the right people at the right time.
This will become especially important as the velocity and advancement of technology create scenarios in which the L1 personnel—followed by L2s and L3s—become overwhelmed with new requests. As Koopmans and Casanova both noted, the efficient routing of messages is just the first step. Where viable, “upskilling” L1 personnel to ensure they can take the load off higher-level operators will also be essential, and AIOps platforms are geared to help companies do just that.
“The pace of change has gotten confusing, even to the people who are making [the technology],” Maguire said.
Making the move to simpler AI-enriched action
Another point Koopmans posits in the keynote: In some ways, the public perception of AIOps today is like the public perception of big data some 10 years ago.
Like big data,” he points out, people know AIOps is bound to be the next big thing, but may not fully understand what it is. And just like big data ten years ago, AIOps’s next big step is to find a mix between low user complexity on the frontend and high backend complexity.
Key here are the “addictive” qualities AIOps platforms offer users—namely tools that make various IT Operations easier for humans. As Casanova says in the keynote, “enhanced human judgment and AI-enriched actions.”
“It’s not just the technology doing something; it’s an assistance of a human action,” he said, adding an example: “If you think about the workflow systems, that’s all full automation. You used to send an email and it got routed […] the last thing we want is a ticket routing that we manually have to move from desk to desk.”
To that end, panelists all appeared to agree that one of the next big strides for AIOps platforms—like big data ten years ago—is providing information and general value in a more simplified way.
Simpler offerings mean higher adoption—including among smaller businesses, which may otherwise lack the resources to take on an outwardly daunting new tool, Noeth argued. And furthering the point, Maguire noted that “only the user interface” of future AIOps programs would grow simpler over time—because higher ease of use on the frontend, coupled with more technical heft on the back, is how successful technologies tend to progress.
Evolving via “the little things”
In one of his closing remarks, Koopmans made the broad point that truly best-in-breed AIOps solutions evolve by making big strides via minor adjustments. In an IT Ops environment, this could be something as “small” as a tool that notes who fixed a problem the last several times and recommends sending similar issues to the same team. The high-level goal is to remove noise on one end and increase quality of information provided on the other—a stacking effect that leaves IT personnel with a lot more time for human tasks.
“I kind of view AIOps as more of a practice,” Casanova said towards the end of the keynote. “It is going to transform how some of these IT Operations are running today.”