AI: the 4th Generation of Process Automation in ITSM (Part 1)

Over the next few days we will be publishing a series of short articles written by our VP of Products, Peter Schneider. This is Part 1 of the series, we will be adding links to the bottom of the page as the other parts are released.

Robotic Process Automation

Robotic Process Automation (RPA) is a new way to assist ITSM tool users in their daily tasks of completing routine tasks. Learning the typical choices on the Graphical User Interface for different use cases, the Artificial Intelligence learns to “mimic” the user’s actions and executes them on behalf of the user.

RPA for ITSM tools might focus in first implementations on automating actions that can be predicted with high certainty such as assigning issues to a person, changing a status based on other changes in the incident, creating a new standard change and so on. However, many if not all of these easy-to-predict actions are already today automated with workflows and scripts that run in the background. The problem-solving skills of support agents for IT issues will be hard to imitate by Artificial Intelligence due to the limited analytical capabilities for years to come.

AI-Assisted Issue Resolution

Artificial Intelligence will be very useful for ITSM tools to help increasing the speed and accuracy of incident resolution. The burden of too much information and badly managed knowledge bases limits often the service desk’s efficiency to recognize similar incidents and their successful solutions. Providing assistance to IT support persons when analyzing, categorizing, and solving issues will be an essential benefit of AI in ITSM.

Machine learning algorithm will be able to identity similar incidents based on scoring mechanism using natural language processing, supervised training as well indexed search methods. ITSM tools will provide a list of similar incidents and their solutions at a glance to the support person when solving an issue in the future. When support persons score the results provided by the AI engine based on their usefulness, then the machine learning will learn even without dedicated trainers. Otherwise senior support persons must train the machine learning engine at regular intervals to ensure a high accuracy of finding similar incidents and their solutions.

Machine Learning for Visual Workflow Automation

Machine learning will make its way to visual workflow automation of ITSM tools dramatically improving the capabilities of workflow engines. Business process automation using workflows with embedded AI capabilities will provide a rich yet flexible toolset for organizations.

Today, IT processes are automated based on business requirements through workflow automation. Many ITSM tool vendors offer code-less automation capabilities by combining visual building blocks to workflow sequences executing tasks, sending out notifications, implementing approvals, and waiting for certain events to occur before moving on.

Future workflow engines will have workflow activities that make decisions based on artificial intelligence. Complex combinations of If/Then conditions in workflows will be replaced with dynamically-adjusted, score-based decisions that continuously learn from past decisions.

Workflow activities will be able to identify the intent of customers using natural language processing and fork the workflow sequence according to a scoring process. The workflow will continue based on whether the intent has been clearly identified, whether the intent has been narrowed down to few known options, or whether the intent is not understood. The code-less design of such workflows will allow organizations to build flexible processes that can be trained over time.

This new way of workflow automation will create more flexible decision making capabilities than the current static process engines. Nevertheless, training of such workflow activities to learn from the past and to understand different intents and their corresponding actions will require more continuous training by business process experts.

We will also be hosting a webinar on Tuesday the 27th of November, where Peter will be discussing the topic of AI with one of our key partners, You can find out more about the webinar and register from the link below.

 View the webinar recording

Other articles in this series

Overview, Part 2, Part 3, Part 4

Peter Schneider

Written by Peter Schneider

I am Chief Product Officer @Efecte. Responsible for product management, product marketing and product strategy.