Cross training in call centers with uncertain arrivals and global service level agreements

We consider cross-training in inbound call centers with non-stationary, uncertain arrival rates and global Service Level Agreements. We investigate the option of cross training a subset of agents so that they may serve calls from two separate queues, a process we refer to as partial pooling. We develop a simulation-based search heuristic that finds near-optimal schedules for a pool of two different queue types. We analyze the benefits of partial pooling and characterize the conditions under which pooling is most beneficial. We find that cross training a modest portion of the staff yields significant benefits even when cross training is costly.