Virtual agents like Siri, Alexa and Google Assistant are becoming mainstream for consumers. Soon, they will be for business users as well.
Two key questions remain unanswered:
- How can IT be prepared?
- Where do virtual agents add value in the IT service delivery lifecycle?
According to the recently released Gartner Hype Cycle for Emerging Technologies, 2017, we are past the “peak of inflated expectations” for machine learning and nearing it for virtual assistants. Next stop, the “Trough of Disillusionment”.
Lots of hype… but where’s the business value?
The concept of virtual agents is intriguing but most provided limited value because they’re restricted to rigid answers to questions asked in specific ways. Hard-coded questions/answers aren’t much better than search portals. What seems like a natural conversation, across disparate topics, is often highly scripted.
The most effective AI-based virtual agents leverage NLP and machine learning (ML) to engage in far more dynamic conversations. They also continuously improve their knowledge and accuracy. Even so, building virtual agents that understand intent, ask domain specific questions and learn in real time requires domain experts and computational linguists to construct dialogs that are relevant to the targeted use case.
When properly developed, AI-based virtual agents can enhance incident management efficiency and become essential resources for every employee.
Incident management is a start, not an end
According to ITIL v3, the IT service lifecycle model focuses on five core areas: Service strategy, Service design, Service transition, Service operation and Continual service improvement.
The IT service delivery lifecycle goes beyond responding to user inquiry, to include processes that require advanced reasoning and decision making skills. Thus, even the best trained virtual agents can’t replace humans IT agents and managers.
Likewise, virtual agents must have the ability to accurately route incidents to human IT agents based on criteria such as incident category and staff availability, when needed. And when incidents are escalated to live agents, they should include relevant information such as related user requests, incident records, change records and knowledge articles. This helps IT comply with SLAs and improve operational metrics, such as MTTR.
Service managers seek strategic benefits from AI-based service management solutions. They want the ability to quickly discover valuable information such as the health of IT operations based on the number of incidents, problems and changes related to specific business systems and services. This information allows managers to proactively assess, design and transition IT resources, vendor relationships and budgets. Also essential to managers is the ability to gain insights into how the AI solution assists human agents and is enhancing ITSM processes, reducing support costs and improving customer sat.
Virtual agent technology is advancing rapidly. However, when leveraged as a disconnected/point tool, as opposed to a unified solution, a virtual agent often falls short of expectations. A holistic approach to AI-based service management helps IT organizations optimize their service delivery strategies and avoid the “Trough of Disillusionment”.
Latest posts by Robert Young (see all)
- Applying AI Technology to ITSM: The Difference Between NLP & NLU (Part Three) - February 20, 2018
- Applying AI Technology to ITSM: A Holistic Approach – Part Two - January 23, 2018
- Applying AI Technology to ITSM: What You Need to Know – Part One - January 15, 2018