This course offers a unique opportunity to delve into the practical applications of data science beyond ad-click predictions, focusing on its potential for social impact. Specifically, it presents a case study derived from a machine learning competition hosted by DrivenData, centered around the challenging task of school district budgeting. By developing a model capable of automatically classifying items within a school's budget, this course aims to facilitate efficient and effective comparisons of spending between different educational institutions. Throughout the course, participants will embark on a step-by-step journey, commencing with the construction of a baseline model that serves as an initial, rudimentary approach. Notably, this involves employing natural language processing techniques to preprocess the budget data for subsequent modeling. Subsequently, participants will have the opportunity to explore and implement their own techniques, comparing their results with those achieved by other competition participants. Finally, the course concludes by examining the winning model, which successfully amalgamated a range of expert techniques to achieve the highest level of accuracy. By following this comprehensive curriculum, participants will gain valuable insights into the practical implementation of data science in the realm of school district budgeting, equipping them with the necessary skills to make a meaningful social impact.