Auto Identification of Problems and Change Requests in ServiceNow

Incident, Problem, and Change Request are primary features of any ITSM company. The daily load of incidents for large companies will be more than 100,000 records. It is really difficult for support staff to go through all incidents manually and suggest related problems and change request to the user also this process is prone to more human errors so in order to address this issue and use support staff time effectively we have created a machine learning model which can automate this process. The proposed model first train the clustering model and similarity model then uses a combination of these models to form the sub-clusters within a cluster which effectively help in routing incident to appropriate sub-cluster and recommend the corresponding problems and change request of that sub-cluster. Reports are also built on clusters and sub-clusters which helps support staff and business people to effectively monitor and maintain the incidents and also helps in taking appropriate actions like creating new problem or change request for a set of incidents in a single go.