Diabetes is a chronic disease that causes billions of dollars in economic losses every year. It consists of two main types: type 1 and type 2. Type 2 diabetes is the project's focus because it is inherently preventable. The cause of type 1 diabetes is not yet known, and there is not a proven way to prevent it [3]. 95 percent of diabetes cases are type 2. We aim to identify county demographic factors that correlate with diabetes rates and illustrate this with a program that is capable of accurately predicting a community's diabetes rate, given the appropriate data. First, we analyzed variables by making scatter plots using Python to extract .csv data. The variables that showed a correlation with the diabetes rates were inputted into our neural network (DANN). A neural network is an effective method to analyze variables because diabetes rates are complex, and the result of many different factors, our team needed a non-linear method to handle the demographic data properly. Mellitus is our project's central creation. It can accurately model a town, its demographics, and its hypothetical diabetes rate. Its vibrant interface is built-in Netlogo 6.1.1. An essential element to its performance is the implementation of DANN with the py extension. Mellitus and DANN can work together to make calculations accurately and effectively exhibit the results. Through this method, we identified five main variables that significantly impact diabetes rates: percent of American Indian and Alaska Native, poverty, education, commute time, and health insurance. Mellitus has proved to be able to consistently predict accurate diabetes rates for New Mexico counties using data from these variables.