Artificial intelligence (AI) technology is serving an important role in IT service management (ITSM) as organisations seek to become more efficient and find ways to avoid repetitive tasks.
According to a survey of 529 IT professionals in March, conducted by SpiceWorks, 40% of large businesses plan to implement intelligent virtual assistants over the next year, while 53% currently use AI in their IT departments.
IT could be a main beneficiary of AI and intelligent automation, with IT professionals estimating that almost a fifth of their current daily tasks could be automated by such technology. The vast majority of respondents (76%) said AI and intelligent automation would help to automate repetitive processes in IT.
Doing more with less
IT teams are under increased pressure because budgets continue to drop, while both technology use and customer expectations of service and support are increasing, says Stephen Mann, principal and content director at ITSM.tools.
Business users are asking IT to take on more work, he says, adding: “There are many frustrations that relate to what now seems age-old mantra of having to do more with less. IT, and potentially the business, is also more complex.”
Mann believes artificial intelligence technology and chatbots can address some of these issues by offering 24x7 support and a quicker response time. This could result in a better customer experience and reduce the number of requests to the IT helpdesk.
Take password resets, for instance. Such requests account for 20-50% of service desks calls, according to analyst Gartner. The manual process kicks off when a user calls the service desk. Staff on the helpdesk then authenticate the user by asking a series of security questions. Mann says the helpdesk agent may need to create a temporary password for the user, who will need to change this to a new password of their choice once they’ve logged in using the temporary password.
The SpiceWorks survey shows that IT professionals believe automating these simple processes would free up a considerable amount of their time.
AutomationEdge has developed just such an automation tool, which requires users to click on the AutomationEdge Password Reset app in the self-service helpdesk portal to initiate a password reset. Users can reset their passwords from the service desk portal or request a temporary password, which is automatically sent to their registered mobile phone. The system can also be used by a colleague to reset a password on behalf of a user.
Keeping pace with DevOps
IT teams can also use AI technology to keep up to date with new services as they are added to the IT portfolio. “One of the biggest challenges in IT is the velocity with which services and technologies are onboarded and changed,” says Forrester principal analyst Charles Betz.
In the age of DevOps and software as a service (SaaS), the time between a bug being identified and being fixed has shortened. “This means keeping up to date on issues and their current resolution is increasingly challenging, and it is in solving this problem that I believe AI will add the most value,” says Betz.
Automating frequently asked helpdesk questions is another easy win for IT. Kaleo Software offers machine learning to categorise similar helpdesk tickets and present a suitable reply. One of Kaleo’s reference customers is US food company Venture Foods, which uses the product to automatically answer “how to” questions for users of its JD Edwards enterprise resource planning (ERP) system.
A growing application of AI in IT operations (AIOps) is the use of machine learning to automate IT admin tasks. AI is used to analyse database and system logs to predict potential issues.
“In the case of AIOps, AI makes it easier for IT operations to handle the increasing number of technology-related alerts,” says ITSM.tools’ Mann. “It removes manual activities, delays and costs through the use of algorithms to automatically cluster and reduce event management alerts.”
He says the use cases for AI technology enable IT departments to deliver better, faster and cheaper services.
Chatbots on the IT helpdesk
Mann says chatbot technology has performed well in the consumer sector, and therefore it makes sense to bring it to the IT department.
“It depends what the service desk is trying to achieve across ‘better, faster, cheaper’. Ideally, the driver should be delivering a better customer experience through quicker resolutions and perhaps offering out-of-hours support,” he says.
Organisations can learn from consumer-facing chatbots, using such use cases to implement similar technology for the service desk. Online users could start by raising a concern through a virtual assistant, which could eventually be passed on to a human agent. If a staff member has to spend time looking at the context of the conversation or asking the user what their issue is, and potentially passing the request on to another member of staff, that creates a negative experience for the customer. Organisations can learn to implement a smoother transition between the AI interface and the human agent in their chatbot service for employees.
Companies need to have a clear plan for chatbots, with implementation “based first on defining the measure outcomes desired: call deflection or left shifting, improved MTTR [mean time to repair], and so on”, says Forrester’s Betz.
“When implementing a chatbot and building the models, it’s important to take as much data as possible, preferably in-channel, like digital data for digital channels. Web chat is a great source, as it replicates the mode of communication more effectively,” he adds. “Chatbot vendors may have existing intent models, which would include cases like password reset, employee onboarding and troubleshooting, from which companies deploying can leverage and build from.”
He recommends companies test the platform with a small team before growing the number of people who can access it. “There’s a certain threshold the chatbot needs to hit when it comes to knowledge and ability to respond to questions with the right answers and content. Users will have less confidence in that chatbot if they try to ask a question and it can’t understand the question.”
ServiceNow is one company that is adding chatbots to helpdesk software. In May, it announced the Virtual Agent chatbot.
According to ServiceNow, Virtual Agent is able to communicate in context with staff. For example, if a user wants a new phone, the chatbot would already know the user’s previous device and network. Virtual Agent can be used across various departments, such as human resources, to complete tasks such as resetting passwords or submitting requests to book annual leave.
ServiceNow says the platform is expected to reduce the number of everyday enquiries made to the IT team by 15-20%.
ServiceNow has also acquired Parlo, a startup which offers chatbots that support natural language queries. ServiceNow says the deal will result in better efficiency with its chatbot product and improved understanding of helpdesk requests.
Going forward with AI
One of the areas where experts believe AI technology could be applied is in scenario planning and improving the decision-making process. AI could also be used to augment IT problem resolution.
“The opportunity for AI is massive – how quickly IT departments can succeed with it is not only dependent on the evolution of AI capabilities but also the ability to deliver AI-enabled services that people want and will use,” says Mann.