Skip to content

How to manage incidents with Machine Learning

Share on twitter
Share on linkedin
Share on email
Share on whatsapp
VISION 17 Congress

Óscar González Campos, Product Manager of Future Space participated in the Vision 17 Congress organized by itSMF Spain, where the focus was on the transformation of Information Technologies to make companies agile and digital.

Generally, Incident Management services have a dubious reputation within companies. However, from an internal point of view, the vast majority of incidents are resolved or escalated in a timely manner once they are reported to the appropriate resolution team. Therefore, it is possibly the route followed by the incident that generates this degradation in the employee's experience.

Óscar González presented a success story within his company, where the need to solve the incident in an agile way and always having the employee as the main character, led his team to ask themselves... Why don't we let the machines learn and suggest solutions to us? In this way, the solution was to implement an automatic learning system. An improvement of between 15-20% in the allocation has been measured.

Assignments are made by different means: self-service telephone service... but regardless of the means, all of them are made on fixed photos, that is, on a cataloguing system established in an instant of time, a training given in an instant of time.

However, companies are living ecosystems, where there is a daily evolution, and therefore they have to sustain improvements in efficiency through living systems capable of learning, improving in a dynamic way, and possibly most importantly capable of distributing knowledge.

The system is based on the same set of inputs as the current systems, thanks to the global knowledge of previous resolutions and their feedback, it is able to present the user with a limited set of options together with the reliability of each of them, meeting the objective of increasing efficiency in IT Incident Management through automatic learning.

To conclude, Oscar left a reflection that is worth keeping in mind: "Within Machine Learning, machines will not replace people, the disambiguation part will remain a human task"

Share the article

Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on email
Email
Share on whatsapp
WhatsApp

A new generation of technological services and products for our customers