As a member of the Wake Forest Engineering research team, I conducted an economic analysis of incorporating solar energy projects onto rooftops of underserved schools in Winston-Salem, NC. I began by calculating the hourly power output of a solar array using sun angle, weather data, module orientation, efficiency, and other key parameters. I then integrated this with a PyTorch-based neural network that modeled power consumption data from 149 North Carolina schools based on building square footage, allowing us to assess hourly power surpluses and deficits throughout the year. Using this data, I developed a financial model in MATLAB that incorporated Duke Energy’s solar buyback rates and energy costs to calculate yearly savings, which I then optimized to determine the ideal solar array size given available federal, state, and private rebates. Our findings were presented in a detailed report, a formal presentation to university officials and environmental advocates, and an interactive ArcGIS website.