Juniper, ITSM, AI, artificial intelligence

Artificial intelligence (AI) and machine learning (ML) are transforming IT service management (ITSM) by helping automate routine tasks, predict system failures and improve incident response times.

However, along with benefits, these technologies are likely to bring with it challenges as organizations look to adopt and integrate them quickly to stay ahead of the competition.

Ivanti CIO Bob Grazioli explained that within the ITSM sphere, AI and automation technologies can drastically reduce ITโ€™s workload.

“By automating routine and repetitive tasks, AI and machine learning not only free up ITโ€™s time to focus on more complex issues and strategic initiatives but ensures that these mundane tasks are performed quickly and accurately,” he said.

For example, in ticket classification, AI can quickly sort through many tickets, which are usually in a waiting queue for days.

“AI can route them to the right service categories or even provide the service itself,” Grazioli said. “This can reduce the wait time from days to hours.”

He added that with the tighter integration between observability tools and service management, incident management is becoming more predicative and reactive.

“With this merger, flagged notifications of events occurring within a service can now be directed to service management, reducing the time it takes to become aware of issues,” he explained.

This enables service management to identify problems more quickly and either resolve them or contact potentially affected customers without delay.

Thoughtful Integration, Planning

Ashwin Ram Ragupathi, ITSM evangelist at ManageEngine, said effective integration of AI and machine learning technologies into existing ITSM frameworks requires thoughtful planning and execution to minimize operational disruptions.

“IT teams can achieve this by assessing how these technologies impact various teams across the organizations,” he said.

For example, AI integrations should start with simple tasks that burden IT teams and address issues with pilot projects designed to reduce manual workloads and boost efficiency.

From his perspective, implementing a gradual approach allows for thorough testing and refinement, ensuring minimal disruption before full-scale deployment.

“IT teams must ensure that AI is compatible and interoperable with existing systems to maintain operational continuity,” he said.

Additionally, clear communication and comprehensive training programs for IT and non-technical staff are vital for AI implementation without disruptions.

Grazioli said it’s important that first an organization figures out the right use cases and secondly, the business understands the data that applies to those use cases.

“This is really important in order for you to develop the right integration of your AI into your processes and tooling,” he said.

He added having that work done upfront makes the integration of AI and machine learning much easier and prone to greater accuracy.

Skills Gaps, Data Quality Among the Integration Challenges

Ragupathi said adopting AI and machine learning in ITSM can present challenges such as data quality and availability, integration with existing systems and skill gaps among staff.

To overcome these, organizations should invest in data cleaning and augmentation processes to ensure high-quality data for training models.

“Conducting comprehensive impact assessments and pilot projects can help mitigate potential integration issues,” he explained.

To bridge skill gaps, providing targeted training and upskilling programs is essential, not just for organizational proficiency but also for remaining competitive in today’s business environment.

“Additionally, partnering with AI experts and vendors can smooth over any AI adoption challenges,” he said. “Addressing these challenges ensures more seamless and effective experiences implementing AI and machine learning.”

Ragupathi noted most ITSM tools already leverage ML capabilities to learn from the historic ticket data for right ticket categorization and technician assignment, as opposed to just sticking with manual or limited rule-based automation.

“This ensures faster and efficient service delivery and ticket resolution,” he explained.

When it comes to predicting system failures and enhancing incident response times in ITSM, ML algorithms can intelligently convert alerts from network and application monitoring systems.

They are then processed into tickets in ITSM tools proactively and trigger appropriate incident responses.

“The resolution processes are kicked in immediately without waiting for manual intervention. This proactive approach ensures minimum business downtime and disruption for end users,” Ragupathi said.

From Grazioli’s perspective, the largest challenge in AI adoption is AI anxiety among employees.

“AI is closing gaps in tool unification and increasing productivity in processes, which can be intimidating to those employees who think AI encroaches on their jobs,” he said. “AI adoption can seem daunting, but it is meant to help employees, not replace them.”

He explained business leaders need to be clear and transparent with workers on how they plan to implement AI so that they retain talented employees โ€“ because reliable AI requires human oversight.

Maximizing Benefits, Minimizing Risks

To maximize benefits and mitigate risks in AI and machine learning adoption, organizations can start by establishing clear objectives and KPIs that provide a framework for measuring success and guiding additional efforts.

Ragupathi recommended regular monitoring and evaluation of adoption processes to allow IT teams to identify areas for improvement while promptly addressing issues.

Additionally, implementing robust security measures to protect sensitive data is critical to mitigating risks associated with AI and machine learning solutions.

“Most importantly, however, it’s crucial to establish a culture of openness and innovation to ensure successful AI and machine learning adoption,” he said.

Grazioli added that before implementing AI, organizations must begin with AI governance and guardrails in the legal department to ensure the ethical, responsible and effective use of AI.

“Once these guardrails are in place, you can then address whether the right use cases are being addressed,” he said.

Techstrong TV

Click full-screen to enable volume control
Watch latest episodes and shows

Tech Field Day Extra at Cisco Live EMEA

SHARE THIS STORY

RELATED STORIES