Case Study—AdWords to AdWards: AI as a Coding Partner in Teaching Online Optimization Algorithms

This case illustrates teaching online optimization algorithms for sequential decision making using generative artificial intelligence (GenAI) as an assistant, making complex concepts in algorithm design, implementation, and evaluation accessible even to nontechnical student audiences. Students are asked to translate an unfamiliar healthcare setting into the Google AdWords problem. The well-known structure of the Google AdWords problem both facilitates students’ conceptual learning and makes it easier for GenAI tools to assist with algorithm implementation. However, the complexity of the healthcare environment makes the mapping difficult, requires discussions with “experts” with domain knowledge and various stakeholders in the process, and raises ethical questions. The case is based on an actual process improvement project aimed at maximizing the use of a hospital’s capacity. The main character, Dr. Kelly, heads an observation unit and had tried a predictive model for patient assignments to her unit versus the hospital wards, which, however, can sometimes lead to inefficient bed capacity management. She is looking to improve patient allocation decisions made in real time. The case empowers students to (1) identify situations in which sequential decision-making models and online algorithms are appropriate and preferable to predictive models, (2) discover the concept of inventory balancing, (3) map its application in an unfamiliar setting (healthcare) to a well-studied setting (advertising) through inquiry and stakeholder analysis, and (4) embrace GenAI as a partner in algorithm implementation and testing. Realistic data, code, and models are provided. The rich context of the case makes it appropriate for use in courses in operations management, business analytics, marketing, machine learning, and operations research, both at the advanced undergraduate level and at the masters/MBA level. Supplemental Material: The Teaching Note is available at https://www.informs.org/Publications/Subscribe/Access-Restricted-Materials .