The coronavirus disease 2019 dominated and augmented all aspects of life beginning in early 2020 worldwide. Research and data generation related to this disease developed alongside its spread. Specifically, a group from Johns Hopkins University made geocoded confirmed cases, deaths, and recovery data freely available through GitHub with daily updates. With this data and an array of appropriate predictors, we developed a Bayesian spatio-temporal Poisson disease mapping model for estimating real-time characteristics of the coronavirus disease in the United States. These models incorporated spatial aspects through the use of Health Regions, temporal aspects through the daily data updates, and community aspects through aggregated data on important population characteristics including information about state issued stay at home orders, percent of the population that is unemployed, and percent of the population that is age 65 and older. Alongside model development, we also created several dashboards for visualization of the statistical model as well as simpler spatial and temporal representations of the disease. These visualizations gave rise to a range of results. First, estimates depend on the region of the US as well as the day of the outbreak considered. Generally speaking, risk of new cases is still higher than prior to issuing of stay at home orders, more so for areas with no or partial stay at home orders. We also see higher risk of new cases in areas with more percentage of older population. In conclusion, there is still much unknown surrounding this disease but statistical models of this caliber have the ability to tease out important relationships with disease spread and population-level characteristics.