SLA Breach Prediction for Incident Tickets
No surprise that in the complex systems run by telecom operators a large variety of errors may occur, causing service outages. To some extent everybody accepts that and the tolerance levels are fixed in SLA contracts. Breaching an SLA is a major business problem. Assessing the severity of the numerous tickets is an everyday challenge for the resolution team leaders.
The presentation demonstrates how a random forest based R solution is able to help the ticket prioritisation and reduce the number of SLA breaches by 40% or more by pointing to problematic cases shortly after the ticket was created. The problem situation, the way of model creation and the outcomes will be highlighted. As contrast, the performance of a similar model built for internal HR services will be mentioned.
Advanced Analytics Junior Manager, Vodafone Group BI
Tamás Molnár is working as the team leader of the data science team in Vodafone Shared Service Center Budapest. As such, he is interested not only in building models, but also the needed resource investment, the practicality of the results and finding possible use cases in the enormous Vodafone organisation. Earlier Tamás held positions at GE, Hungarian Telekom, Raiffeisen Bank and IFUA Horváth & Partners as data scientist and BI consultant. Has a total of 10 years experience in data mining besides other BI experience.