IT Service Management (ITSM) has become even more efficient with machine learning tools.

In 2018, we firmly established our position - the IT Services Management Services (ITSM) and IT services continue to continue their activities, despite the incessant talk about how long they will last in the digital revolution. Indeed, the demand for technical support services is growing - in the Technical Support Report and the HDI Salary Report (Help Desk Institute) for 2017, it is indicated that 55% of technical support services noted an increase in the volume of applications for the last year.



On the other hand, many companies have noted a decrease in the volume of calls to technical support in the past year (15%) compared with 2016 (10%). A key factor contributing to a decrease in the number of applications was independent technical support. However, HDI also reports that last year the cost of the application increased to $ 25, compared with $ 18 in 2016. This is not what most IT services aspire to. Fortunately, automation based on analytics and machine learning can improve the processes and productivity of the help desk by reducing errors and improving quality and speed. Sometimes this goes beyond human capabilities, and machine learning and analytics are the key basis for intelligent, responsive and responsive IT support services.

This article discusses in more detail how machine learning can solve many problems of support services and ITSM related to the volume and cost of applications, and how to create a faster and more automated support service that enterprise employees will be happy to use.
')

Effective ITSM through machine learning and analytics

My favorite definition of machine learning is provided by MathWorks :

“Machine learning teaches computers to do what it is natural for people and animals to learn from their own experiences. Machine learning algorithms use computational methods to study information directly from data, without relying on a predetermined equation as a model. Algorithms adaptively improve their own efficiency as the number of samples available for study increases. ”
The following capabilities are available for some ITSM tools based on machine learning technologies and big data analytics:

• Support through the bot. Virtual agents and chat bots can automatically offer news, articles, services, and support offers from data catalogs and public requests. Such 24/7 support in the form of proposed training programs for end-users helps to solve issues much faster. The key benefits of the bot are an improved user interface and fewer inbound hits.
• Smart news and notifications. These tools allow you to notify users in advance of potential problems. In addition, IT professionals can recommend workarounds to solve problems with personalized notifications that provide end users with up-to-date and helpful information about the problems they may encounter, as well as tips on how to avoid them. Informed users will appreciate active IT support, and the number of incoming calls will decrease.
• Smart search. When end users search for information or services, a context-sensitive knowledge management system can provide recommendations, articles, and links. End users usually skip part of the results, preferring others. These clicks and the number of views are included in the "weighting" criteria when content is re-indexed over time, so search capabilities are dynamically tuned. Since end users provide feedback in the form of a “like / dislike” vote, this also affects the rating of content that they and other users can find. In terms of benefits, end users can quickly find answers and feel quite confident, and help desk agents have the ability to process more requests and reach more service quality agreements (SLAs).
• Analytics of popular topics. Here, analytic capabilities reveal patterns for structured and unstructured data sources. Information about popular topics is graphically displayed in the form of a heatmap, where the size of the segments corresponds to the frequency of certain topics or groups of keywords demanded by users. Repeated incidents will be detected instantly, grouped and resolved together. Analytics of popular topics also detect incident clusters with a common root cause and significantly reduce the time to identify and solve the underlying problem. The technology can also automatically create knowledge base articles based on similar interactions or similar problems. Finding trends in any data increases the activity of the IT department, prevents the recurrence of incidents and, therefore, increases end-user satisfaction while reducing IT costs.
• Smart application. End users expect that sending a request is no more difficult than writing a tweet, namely, a short message in natural language describing the problem or request that can be sent via email. Or even simply attach a photo of the problem and send it from your mobile device. Registration of a smart application speeds up the process of creating a request by automatically filling in all fields based on what the end user has written or scanning an image processed using the optical character recognition program (OCR). Using a set of observations, the technology automatically classifies and addresses applications to the appropriate support agents. Agents can redirect requests to various support groups and can overwrite automatically completed fields if the machine learning model is not optimal for this case. The system studies on new patterns, which allows to better cope with emerging issues in the future. All this means that end users can easily and quickly open applications, which leads to increased satisfaction with the use of working tools. This feature also reduces manual work and errors, and helps reduce the time and cost of resolution.
• Smart email. This tool resembles a smart application. The end user can send an email to customer support and describe the problem in natural language. The support tool creates an application based on the email content and also automatically responds to the end user with links to the proposed solutions. End users are satisfied, because opening applications and requests is easy and convenient, while IT agents have less manual work.
• Smart change management. Machine learning also supports modern analytics and change management. Given the frequent number of changes that enterprises currently require, intelligent systems can provide agents or change managers with proposals aimed at optimizing the environment and increasing the percentage of successful changes in the future. Agents can describe the required changes in natural language, and analytical capabilities will check the content for the presence of affected configuration elements. All changes are regulated, and automatic indicators tell the change manager whether there are any problems with the change, such as risk, planning in an unplanned window, or “not approved” status. The key advantage of smart change management is a faster payback time with fewer configurations, settings and, ultimately, lower cash costs.

Ultimately, machine learning and analytics will transform ITSM systems using intelligent assumptions and recommendations about application problems and the change process that help IT agents and IT support teams describe, diagnose, predict and prescribe what happened, what happens and what will happen. End users receive forward-looking, personal and dynamic analytical evaluations and quick decisions. In this case, much is done automatically, i.e. without human intervention. And as technology learns over time, processes only get better. It is important to note that all the intellectual functions described in this article are available today.