An Agent Scheduling Optimization for Call Centers

A suitable assignment of agents to handle the workload at a call center is not easy to fulfill because it requires a combinatorial optimization with large solution space and the number of incoming calls is subject to sharp fluctuations over the course of the day. At most call centers, the agent scheduling is done manually using spreadsheets or other software; this takes a lot of time and effort, but the results are often that the center is overstaffed or understaffed. We propose an efficient and practical method to determine the schedule by using traffic theory and evolutionary computation, which has not been applied before to this kind of combinatorial optimization. Heuristic restrictions are introduced for the initial parameters and the search procedure, which had been given randomly in preceding researches. It is shown that the rapid calculation and high accurate scheduling are achieved even though the evaluating function is simplified.