Abstract
Service-delivery modeling and optimization is a complex problem involving multiple service levels, skill sets, request classes, and service times. In this paper, we describe a comprehensive and scalable end-to-end analytical methodology that we developed and implemented at a global information technology service delivery provider. This methodology provides predictive insight and prescriptive solutions to the problem of staffing servicedelivery units in this complex environment. Our solution has been deployed globally at more than 640 servicedelivery units and has yielded more than $52 million in cost savings and cost avoidance to date.