In real life, supply and demand needs play an important role in the opera-tion of queuing systems. Queuing systems are expected to be well designed in terms of energy and time savings. In this study, an agent-based queuing system model is defined. This study aims to obtain the optimum result by considering the performance criteria of the queuing systems. The arrival and service times of the agents constituting the system’s population are calculated according to probability distributions. A case study is presented on an inventory problem. The agent-based queuing system model is run by agent servers. The system consists of a finite calling population, the nature of agents’ arrivals, the service mechanism, and the system capacity. The system has a First In First Out (FIFO) queue discipline. A model with a multi-server and multi-channel structure has been defined. A backup agent server is defined as a precaution against possible malfunctions in the system. In this study, parameters such as the average number of agents waiting in the queue, average waiting time, time spent in the system, and service times are observed. Experimental studies are carried out with 20, 30 and 40 agents, respectively. The benefits of the system with single server and multiple server agents are observed depending on the waiting times and service times. According to experimental studies, it is revealed that the optimum result is multi-server systems running at different capacities for a 40-agent system according to the minimum waiting rate and minimum service time results. At the end of this study, the queuing analysis of the agents is performed and the profit-loss balance of the inventory system is discussed comparatively.