AI often seems like a solution looking for a problem, a trap that often snares information technology. But there’s no disputing that AI has transformed IT service management, which is in the business of solving problems. By automating repetitive tasks, AI has moved beyond managing the ticketing system backlog. AI’s next moves will be predictive: injecting AI in service management processes will minimize downtime and improve system stability.
The tech team’s predictive AI journey often starts at the help desk. In March, Gartner predicted AI agents in customer service would cut operational costs 30% by 2029. However, by sharing data across applications, AI agents can also help fix what breaks most often. By quickly running thousands of real-world scenarios, machine learning will find the Achilles heel of a tech environment – the technology or human process most likely to fail or with a pattern of failure. The technology team can then devote resources to optimizing the process.
While teams stem the bleeding, managers can diagnose the root causes and heal the system. AI helps managers spot process bottlenecks, improve tech training, prompt employees for feedback, implement solutions and continuously monitor results. It’s a virtuous cycle that sets clear objectives for the best gain in staff efficiency, cost cutting or customer satisfaction.
Virtualizers gear up for real gains
The internet of things is a lake teeming with information. AI can catch data at the source and trace its many possible paths downstream. On its own, a count of the number of service tickets, or even the ratio of open issues, does not tell people watching the internal tech team dashboard much more than how busy they all are. But what are their most frequent issues? Are they related? Do they recur? Do complaints languish in the call center and become service-level breaches?
AI can analyze complex chains of causation to find a root cause. Furthermore, AI-enhanced automations speed their resolution, showing the team the problem proactively rather than waiting for it to reveal itself. When an AI-enhanced process flags data storage before it’s 90% full, a team can then take steps to prevent a server crash.
Hybrid cloud systems build AI protections into hypervisors – virtual machine monitors that emulate multiple computer systems. By leveraging these cloud virtualizers with on-premises software, companies have more flexibility to allocate their resources. Now, with proactive service management, AI automatically adds inexpensive, stable computing or storage to achieve the best outcome for the customer.
The city of Chicago shows how AI has evolved. When the city was building one of the country’s biggest networks of public outdoor cameras, there was one problem: not enough monitors. While the Chicago Police could install cameras at prime crime hot spots, there would never be enough police officers to monitor hundreds of camera feeds in real time. That was over 15 years ago. Now, object recognition algorithms can spot a briefcase or backpack that is out of place and alert officers in the surveillance room.
In this case, natural language processing has also grown more powerful, and call centers have benefited from this transformation. If an end-user says, “My email is really slow today,” the contact center software automatically categorizes the issue, gauges its severity and identifies potential causes.
Call takers get this real-time analysis, and if they’re using intelligent voice recognition, they can concentrate on the customer and follow AI prompts to resolve the issue. Voice recognition software is also more sensitive to a caller’s tone, a perceptive ability that helps keep users happy.
AI’s KPIs close the customer feedback loop
Chatbots, virtual agents and robotic process automation will increasingly handle routine tasks like ticket routing and password resets. Automation will free the tech team for more challenging, strategic tasks. By empowering end-users and frontline support teams, AI will shift problem-solving closer to the source. This “shift left” approach reduces escalations and resolution times.
Key performance indicators will offer a holistic view of AI and automation’s impact. End-user feedback can reveal how well the technology is meeting expectations. Regular shadowing sessions can help assess how IT service desk agents interact with AI tools and automated processes in the real world. The best outcome may be whether customers spend less time with the call center.
IT organizations can expect greater complexity in the issues that reach human agents, so KPIs should also consider work experience. To gain deeper insights, it’s important to ask employees open-ended questions about how automation has impacted their job satisfaction, stress levels, and opportunities for professional development. Once routine tasks are automated, conducting interviews will help determine whether employees feel empowered to tackle more challenging problems.
The biggest challenge
The biggest challenge in introducing AI into the workflow may be the learning process. An AI rules engine is constantly adapting, and like riding a bike, the process takes patience. Chief information officers will have to take early missteps in stride. However, AI’s advanced performance metrics and deeper insights will align IT strategies with business outcomes, enabling better data-driven decisions.