<p>This study contributes to the literature on unrelated parallel machine scheduling problem by investigating a scenario in which a single machine is required to perform more than one job. This practical restriction plays a significant role in real-world problems; however, previous studies have neglected it. Therefore, this research introduces a different interpretation of job batches and families and formulates a novel mixed integer linear programming model to minimize the makespan, the total completion time of jobs, and the costs of employing machines. To tackle problems of larger size, a hybrid evolutionary method based on the combination of two efficient metaheuristic algorithms is developed and enhanced by a local search. The proposed hybrid algorithm solves a case study in a software development project. The results demonstrate that the proposed algorithm outperforms well-known benchmark algorithms.</p>