As Federal agencies grapple with increasingly complex IT environments, AI Ops – which applies AI to automate IT operations – is a growing approach for organizations to identify and resolve system problems faster, experts from GDIT and ServiceNow said in a recent conversation.
Traditional IT service management (ITSM) focuses on mean time to repair, while AI Ops enables organizations to be more proactive. “AI Ops is able to shift left … mean time to detect. How soon can you find the problem?” explained Arun Iyer, principal enterprise architect and field chief technology officer for ServiceNow.
“Once you find the problem, solving it is rather easy,” he continued. “Mean time to detect is where AI and AI Ops is going to help, because you are now using data from your repositories and log files to determine the patterns. If you know the patterns, can you use a solution that you used before?”
As AI Ops has evolved, experts say, it has become even more proactive and follows a clear progression aimed at ultimately creating self-healing IT systems that not only predict issues but remediate them autonomously – driven by the integration of AI within the IT environment.
Patrick Fredericks, senior solutions director at GDIT, seized on that perspective during the discussion, noting that GDIT customers “want us to shift from being reactive to proactive using the tools we have” to solve IT issues – and that AI Ops can help.
With today’s IT infrastructure spanning physical servers, cloud services, and containerized applications, sophisticated service management approaches are more important than ever. Two-thirds of Federal agencies report embracing emerging technologies to accelerate their digital transformation, according to research from GDIT. Robotic process automation (RPA) and generative AI (GenAI), in particular, are fueling the adoption of AI Ops.
RPA is a crucial bridge between legacy systems and modern cloud-based solutions. “The increased adoption [of RPA comes from] being able to bridge the gap between different tools, technologies, and products across the evolution of the technologies,” Iyer noted. This integration helps overcome what industry professionals call “swivel chair integration,” where staff must manually transfer information between different systems.
“Without RPA,” Iyer said, “you have legacy systems which don’t have APIs (application programming interfaces), and you have newer systems where people are expecting everything to work over the internet, and now you have on-premises data centers, [and] the cloud. In all of this, you’re depending on the individual at the terminal to make the right choices.”
While RPA handles repetitive, rule-based tasks, automating manual processes, GenAI provides intelligence to those tasks, enabling them to adapt to changing conditions. Applied to ITSM, “GenAI provides the information to get better results,” Iyer said. “Rather than having to pore through different documents, you’re getting a summary of the underlying data in a more consumable fashion.”
In the shift to proactive service management, GenAI is invaluable for IT managers needing to sift through growing volumes of data.
“It’s very impractical for humans to sift through all of this data,” Iyer said. “GenAI speeds up decision-making.”
Two critical considerations as agencies employ AI and RPA are data management and decision-making oversight.
“The underlying data needs to be collected in a standard fashion for you to be able to solve problems,” Iyer cautioned.
And human oversight remains crucial. The industry recognizes three stages of human involvement: human in the loop, human on the loop, and human out of the loop. Currently, most organizations are at the “human in the loop” stage, Iyer said, where AI provides decision support rather than making autonomous decisions.
“Human out of the loop is a little far, yet, in my opinion,” he noted.
For Federal and state and local organizations getting started with AI and RPA, Iyer advised identifying their top 10 systems.
“If there’s an outage, which systems do you bring back up first?” Iyer asked. “Do you bring back up a time sheet system or your storefront? Do you bring back a social media server or some services you’re providing to your citizens?”
Once agencies know their 10 most important systems, they can look for opportunities to apply AI and RPA, the experts advised. To learn more about how RPA and GenAI can be applied to ITSM – and how they help fuel AI Ops, view the discussion.