We report on a scheduling app that provides optimal shift assignments for in-home healthcare workers. The problem involves substantial complexity due to various training requirements, legal restrictions, and high variability in both demand and supply with irregular hours. Moreover, in-home healthcare services involve significant face-to-face interactions, and so patients’ and employees’ subjective preferences are a major factor. Most organizations deal with such complexities by determining assignments manually with schedulers; however, this requires substantial training, involves high time cost in developing schedules, and elevated costs due to typically suboptimal solutions. Our solution involves mathematical optimization, deployed via a mixed integer programming model. The model is inspired by a real-world problem of helping to schedule health aides to reduce overtime expenses and achieve equitable schedules in a non-profit organization; this involves 50 weekly shifts across 40 locations and 30 employees. Our pilot test suggests that our software can reduce much of the overtime costs present in manually configured schedules. We also discuss the scheduling app that integrates our optimization model into the existing system.