Modeling and understanding the spread of disease has been a topic of much focus for epidemiological researchers in recent years due to the effects of the COVID-19 pandemic. High levels of global attention and an abundance of recently collected data have created an environment for epidemiological models to be highly detailed and impactful. The best of these models incorporate ideas and research from a range of domains, with clarity and ease of consumption being key focuses so that they have the highest chance of impacting public policy. Here, probabilistic planning is leveraged to understand the spread of COVID-19 at a regional level in the city of Kingston, Ontario, through an agent-based model implemented in RDDL. This data-driven model operates with the functionality of introducing mask and vaccine mandates as sparingly as possible, as per policies created by JaxPlanner, while attempting to ensure that hospitals are operating below capacity. Mimicking the variability of real data, a variety of model configurations are experimented with, and the resulting simulation differences are noted. The observed dynamics are well in line with the theoretical models, and the intervention methods utilized proved to be successful at the task of mitigating hospital burden, though the policies decided upon were inconsistent at times.