The Realities of Implementing a Cognitive Service Desk (Part 2)
Author(s): Paul Gleeson, Posted on March 8th, 2017
In part 1 of this blog series, we introduced the growing interest in deploying artificial intelligence into Service Desk operations and the segmentation of service incidents that would most benefit from automated resolution. In part 2, we address the foundational IT, People and Process elements for successful implementation and adoption of AI-enabled response and resolution at the Service Desk.
Architectural Foundation. For an AI system to support nondeterministic incident resolution, there needs to be a foundation architecture in place – an integrated environment that provides the AI system with the knowledge, data and analytics at the very moment the incident occurs. If this foundation architecture is not there, then the ability of an AI system to resolve nondeterministic incidents is significantly reduced and the effectiveness of the investment is impacted.
In our previous example, in order for the AI system to make the correct decision on the resolution, it will need a significant amount of live data inputs. These inputs range from having the ability to interrogate the end user device to determine if the current version of browser and the plug-in are compatible to accessing network performance data from the cloud provider’s infrastructure to see if there is some sort of issue there. Gathering the data input is only the first phase in resolution. The second phase is to use analytically-driven experience and knowledge-management-supplied resolution paths to determine the best course of action to resolve the incident. The third phase is the actual process of executing the resolution using a combination of automated executable scripts, synthesized voice guidance and directing the user to self-resolution knowledge articles.
These live data inputs, comprehensive knowledge articles and analytics, while very achievable with modern tools and technology, are expensive and complex to implement and, unsurprisingly, often are not in place before commencing with an AI Service Desk project, thus adding to the overall project cost, complexity and duration.
It is worth noting that it is becoming a common AI deployment strategy to use human intervention to deal with the more complex and exceptional nondeterministic resolution incidents and avoid the issue of foundation architecture.
Cultural Foundation. There is an apocryphal quote often attributed Peter Drucker that states, “Culture eats strategy for breakfast.” Regardless of its origins, that assertion conveys an important truism in the corporate world: understanding an organization’s appetite for change can be a leading success factor when implementing a culturally impactful project. This is particularly relevant with Service Desk AI projects. If an organization is culturally not ready to accept working with a synthesized service desk agent, a comprehensive organization change management and end-user adoption plan is required to successfully deploy the technology.
Tell them what you told them. Artificial Intelligence is one of the fastest growing areas of technology in the market today, with exciting use cases being supported by new market entrants. However, as with any hype cycle, some vendors are overstating the benefits of this technology, especially in the area of Level 1 Service Desk. To best judge how an organization can benefit from AI technology, there should first be a deep dive assessment of the profile of the current Level 1 incidents in the environment. This analysis will provide some good insight into amount of incidents that can be adapted to automation. The incidents that are deterministic are significantly easier and quicker to automate that nondeterministic ones, because these incidents require a complex foundation architecture to ensure success. Finally, it is important to assess the cultural appetite of adopting AI technology, because this can be a major contributor to the success of the project.
There is little doubt that AI adoption is going to continue to expand over the next few years as AI engineers get smarter and their efforts more cost effective. As such, organizations’ IT leaders should be preparing a strategy for how this technology can help reduce costs and improve the user experience. Even though they lost, we have to thank Mr. Rutter and Mr. Jennings for this exciting opportunity.
Tags: Applied Service Management, Artificial Intelligence, Automation, Autonomics, Cloud, Cognitive Agent, Digital Business, digital IT, Digital Service Management, Digital Service Support, hyrbid IT, IT Operations, ITOM, ITSM, Service desk