Policy Based Service Management

Policy Based Service Management is where you are able to define your service delivery metrics/parameters within the application such that everything is “automagically” applied throughout the lifecycle of the request. The service desk does not have to manually apply priority or manually escalate tickets/incidents/requests or determine which queue or person it should go to – all key elements of providing leading edge service delivery. If part of your service delivery contains manual decision points such as these how is AI going to help?

From a service delivery process this means when the service desk takes a call they can focus on the issue at hand and quickly identify trends (i.e. potential Major Incidents), they do not need to worry about what priority the request should be set to or what group this may need to be escalated to if it can’t be dealt as an FCR/FLR. Your policies would define the priority based on the type of request, who the end user is, where the caller is calling from (any number of criteria) , the resolution and callback targets if required, the subject matter experts/groups that should be involved and any notifications/messaging that needs to be applied. When this is all orchestrated to align your service desk with your delivery objectives the service desk can focus on the content and the delivery rather than trying to manually (and often incorrectly) align itself with what the business objectives are – these objectives are known quantities that are “programmed” in.

So before we go and jump on the AI band wagon, let’s make sure we have the foundation in place to ensure that the implementation of new technologies has a chance to succeed. IncidentMonitor™ combines AI and policy based management to facilitate your request intake by automatically applying the correct classifications and policies based on the content within the ticket/request. The framework and application should have core capabilities to define your policies and enforce them without manual intervention. Only then can you look at automating the intake through a Turing like bot.