Service Management Artificial Intelligence Use Cases to Implement in 2020

The vision for implementing artificial intelligence in the business is to augment and improve the productivity of existing resources, while improving the customer/user experience. Technology trends and marketing buzzwords seen echoing through conference halls continue to reverberate the excitement and importance of adopting artificial intelligence. In this article, we explore actionable artificial intelligence use cases that I believe service management organizations will continue exploring and implementing in 2020.

Through the adoption of AI technologies, organizations providing service management will be able to provide both faster and more accurate service across the customer journey. AI promises to offer improvements on both sides of the support spectrum (making the end user experience better, as well as improving the technician’s ability to solve issues). When considering Michael Porter’s three generic business strategies of cost leadership, differentiation, and focus, the primary goal for service management providers is to adopt AI to maintain differentiation in the superior innovation, quality, and service of the organization. A secondary effect of implementing AI is driving cost leadership through the automation and optimization of existing processes and tasks.

When looking at the service management artificial intelligence use cases to implement in 2020, we will specifically be focusing on the areas I believe have matured beyond hype and to the point of adoption. Those shorter term use cases are aligned to trends in Machine learning and Natural Language Processing/Generation will transform the way the organization does business.

Supervised and Unsupervised Machine Learning Use Cases

Automatically Routing and Enriching Tickets: When support tickets are opened for service management organizations, these traditionally need to be manually (or automatically via predefined data definitions/rules) triaged to the proper groups for support. Machine learning can improve on the manual error of mis-categorization and assignment by either taking over high confidence categorizations and assignments automatically, or passively recommending the teams and categories to technicians. In addition, machine learning can be used to predict and set the fields used to drive existing data definitions/rules.

Empathy Analysis in Support Calls and Tickets: Calls and tickets, and conversations during them, are handled daily between service management providers and the employees and customers submitting them. Machine learning could be applied to understand user sentiment and inform the staff of changes to approach. Empathy comes from sentiment and tone analysis, which is a combination of supervised machine learning and natural language understanding, and these areas can help provide automated escalation to tickets.

Resolution Recommendations in Tickets: Unsupervised machine learning techniques around clustering are useful to provide insights into similar pieces of data and information. For technicians working on tickets, this can show similar tickets that were resolved, as well as similar content that could prove useful in understanding the resolution steps needed to take to solve an issue.

Visual Recognition: Neural nets and standard OCR algorithms mean that pictures can be analyzed to have their contents and category identified. This allows employees to seek help on topics via simply taking pictures of the problems they are encountering. Technicians can use similar approaches to understand manuals and documentation for hardware and assets they are working with in the real world.

Natural Language Processing and Generation Use Cases:

Virtual Agent with Natural Language Understanding: A virtual agent would facilitate self-service and automated deflection of issues. In addition, chat offers a cheaper issue creation channel than phone. It can answer general questions by searching across multiple content sources, or trigger automated resolutions via integrations with backend systems. Natural language understanding allows for intent identification to properly identify the users issue, with entity extraction to automatically pull out the proper context.

Call & Conversation Summarization: Service provider personnel are frequently on the phone with customers. A speech to text and auto summarization would automate this process and can help surface resolution recommendations when used alongside a speech to text system. When tickets are resolved after dialogue between a technician and the customer, the resolution details are captured to generate knowledge that a future technician could leverage. Auto summarizing the resolution details would facilitate the process.

More Accurate Results Searching and Querying: By automatically tagging documents with meta data pulled from entity extraction and intent extraction, search becomes more accurate. Semantic search alongside knowledge graphs can also help power context by linking relevant concepts in a search space. In addition, the introduction of natural language querying will allow systems to translate text into specific database queries for retrieval.

Multilingual capabilities via machine translation: Another space commonly dominated by neural networks and linguistics is that of machine translation. As globalization brings together customers, employees, and businesses, the need to break language barriers has grown. Machine translation has provided a quick and continually improving alternative to connecting service providers and consumers in their native languages.

Voice capabilities: Additional advances in linguistics, electric engineering (hardware) has allowed for the commercialization of speech to text and text to speech capabilities. Growing acceptance of IoT and smart home trends such as Siri and Alexa have also normalized the concepts of utilizing speech as a method for engaging with a system to find answers.

Concluding Thoughts and Look Ahead

The service management AI use cases to implement in 2020 offer great improvements to those providing service and those receiving it. A majority of the use cases that we observed involved enhancing and simplifying the experience of employees and service providers by augmenting interactions with more data. From Porter’s perspective, cost leadership is a clear benefit of introducing AI as operational expenses are minimized in a support context as virtual agents replace tier 1 labor and allow existing labor to focus on higher complexity, higher value add opportunities. In addition, they are available 24x7, and can potentially support multiple languages offering native scalability. Knowledge can more quickly be documented, and self-service opportunities are available on additional voice based contact channels. Multi-lingual machine translation means cheaper labor can be found, without requiring speakers native to a user’s location. Differentiation is another benefit aligned to Porter’s forces, as the auto generation of summarized text based on voice conversations held by support staff over the phone or in an incident (case ticket) will significantly proliferate the amount of information available to analyze, and to re-use in future interactions. This, alongside automated sentiment analysis help offer a differentiation point of more effective customer service. The automated routing and recommendation of solutions means resolutions times will decrease and organizations can provide a competitive service that is differentiated on faster service.

So what does the future beyond 2020 hold? I predict a greater increase in use cases following the trends of robots, internet of things, and robotic process automation. Robotics will continue to increase in terms of delivery, cleaning, maintenance/repairs, and customer service. IoT will continue to grow in organizations as more sensors are placed for workplace management and proactive ticket management through pattern detection on the collected data. Finally, robotic process automation is a field that will mature as the next flavor of integration/orchestration activities that replaces manual, repetitive tasks and strengthens ‘self-healing’ systems.

While the benefits of AI are strong, there are two primary areas of caution for consideration: data privacy and labor reduction. When looking at auto summarization and speech to text services, there is concern over privacy and over IP generated. When looking at machine learning based recommendations, there is machine learning bias to consider from a biased dataset. Organizations must be explicit on disclosing data collected, and be sure to maintain anonymity of PII and data in aggregation techniques. The reduction of human labor required is a potential side effect of the computationally superior abilities of machine learning in classification settings, as well as the ability for virtual agents to augment or replace tier 1 technicians. Organizations must focus on workforce training to enable lower skilled workers to begin solving higher complexity tasks. They must also prepare new roles to maintain the solutions.

Is your organization already up-taking AI in your service management journey? Let me know which of the use cases and trends above your organization is pursuing, and if there are any not listed above.

Don’t know where to start in your process of driving AI based digital transformation in your organizations processes and employee experiences? Contact a ServiceNow representative today to learn how to take artificial intelligence service management strategy into execution.