Artificial Intelligence

In a recent artificial intelligence (AI)-related HDI webinar, Roy Atkinson of HDI, Stephen Mann of ITSM.tools, and I fielded questions related to what service and support managers need to know about AI. There were some pre-planned questions plus those offered by the audience. This blog shares some of the key AI-adoption points raised and the advice that was offered.

The audience’s position on AI

Before our panel fielded any questions, a poll was undertaken to ascertain where the attendees, and their organizations, are in their AI journey. This showed that:

  • 25% are already using AI in certain applications
  • 25% are currently experimenting with AI
  • 27% are planning to begin using AI in the next 12 months
  • 8% have no plans to use AI within the next 12 months
  • 13% don’t know what their organization’s plans are.

However, as you would expect from the attendees of an AI-related webinar, they’re highly likely to be more interested in AI than the average IT service desk manager.

Other IT service management (ITSM) industry surveys, related to IT service desk AI-adoption levels, have shown that it’s closer to a 50-50 split between those doing/experimenting/planning and those with no plans. But the number of webinar attendees and the above poll are definitely evidence of the growing interest and use of AI-enabled capabilities for IT support (or support services offered by other business functions in enterprise service management scenarios).

  1. What AI is and is not

While the two words “AI” and “automation” might sometimes be used interchangeably, AI is different from automation – with this key in understanding the differences between the two. There are many definitions of both, some dating back decades, but a simple way to differentiate between the two is that:

  • Automation technology follows a specific set of instructions, usually set by a person. The technology is bound by the instructions/rules, with it simply “following orders.”
  • AI technology mimics humans – thinking like a human and acting like a human. And experience-based machine learning technology learns new patterns and outcomes to improve.

Then there’s also “automation and AI” to consider – where the AI technology invokes automated actions. However, IT service desk managers, or any potential adopters of AI capabilities for that matter, need to be wary of products/services that are sold as though they are AI when they’re really automation. For example, scripted chatbots with pre-programmed logic and flows – with this automation, based on the above definition, rather than true AI.

  1. How AI is being built into ITSM tools

There are many opportunities for AI exploitation in the world of service management and support. Some of the key AI-adoption areas already appearing in ITSM tools include are Predictive Analytics Models for IT Service Management, Process and Application Automation, Digital Agents and Conversational User Experience  

 

  • Digital agents for automated resolutions – these use natural language processing to understand an end-user’s need and invoke automated resolutions to immediately resolve or provision against it. But please note that there’s a spectrum of available technologies here – from chatbots that answer questions to more powerful digital agents that can also do what’s needed.
  • Staff augmentation via AI-delivered intelligence – where service desk agent knowledge, capabilities, and productivity are all boosted by what the technology knows and can do.
  • The intelligent automation of high-volume, low-value tasks – a good example is with ticket “catch and dispatch” where the AI will categorize, prioritize, and route tickets to the right resolution group. With this instantaneous versus traditional practices and with no human involvement (or cost) involved.
  • Context-based knowledge provision – where the technology offers up likely solutions, or information, while the end user is using more traditional self-service capabilities.

As to when these capabilities will be available – the quick answer is that they’re already available now https://www.symphonysummit.com/products/summitai/

  1. How next-generation tools will affect the day-to-day operations of service and support

The impact of AI will be felt on the IT service desk in a number of ways. Firstly, and importantly for any service desk agents reading this – the technology is very much focused on replacing low-level tasks rather than replacing roles (or jobs).

Then AI is, and will be, something that allows service desk agents to “work smarter, not harder.” For instance, in removing the many high-volume, low-value tasks from their plate such that they can focus on higher-value-add work that’s also likely to be more rewarding. Or providing knowledge and intelligent/automated capabilities that make working on the IT service desk easier – think of this as driving with a sat-nav versus with a paper map (or no map at all).

There are also knock-on impacts to consider too though. For example, if AI removes many of the simpler tickets from the IT service desk (and its people), then service desk agents will be fielding more complicated issues and requests on average. These will likely take longer to address. Here traditional metrics and targets will need to be reassessed to reflect fewer tickets, increased handling times, and a lower level of first contact resolution (FCR) among other changes to work patterns.

  1. How support organizations should be preparing now

While it’s key to understand “the art of the AI possible” – cutting through the industry hype as needed – and how it applies to your IT service desk needs and potentially the wider organization. There’s also a number of other things to consider up front. For instance, applying many of the same checks to AI capability adoption as you would to any other new technology solution, such as:

  • Ensuring that the business’ needs are put first
  • Not investing in technology “just because you can”
  • Avoiding a disjointed or siloed approach to technology investment and adoption
  • Ensuring that the right people and skills are available – from those who can get the most out of AI solutions to those who will use it on a daily basis
  • Not simply replicating manual processes, and instead seeking to improve operations and outcomes based on the offered capabilities.

Of course, there are other many checks that could be added to this list, but it’s a good start in ensuring that your preparation is focused on the outcomes of AI adoption rather than the technology itself.

Audience question topics

There were also a number of great questions from the audience, including ones that related to:

  • The cost model for AI capabilities and how this differs for software-as-a-service (SaaS)-based offerings versus organizations creating something internally.
  • AI adoption in enterprise service management scenarios or where hosting providers need to drive efficiencies in managing infrastructure events.
  • Whether CINDE (SUMMITAI’s “digital agent”) connects to an ACD, allowing customers to opt out of the AI capabilities and be routed to a service desk agent.
  • Whether AI could help with “next issue avoidance” – addressing and preventing possible follow-on issues after the initial resolution is provided, and thus preventing call-backs.
  • Whether AI tool learning and exploitation can be supported at the service desk instead of needing to hire developers or vendor services for this work.
  • The best ways to use AI to deflect L1 services and to support shift-left initiatives.
  • Advice on integrating knowledge when training AI.

If you would like to hear answers to these questions, and more detailed information on the points outlined above, then please take the opportunity to watch/listen to the on-demand version of the webinar that HDI has made available here.

Finally, if you have any other questions related to how AI could be helping your IT service desk, its staff, and its end users right now, then please leave a comment below or contact me via akhil.sahai@symphonysummit.com

Leave a Reply

Your email address will not be published. Required fields are marked *

Post comment