This course develops the ability to translate a bounded organizational problem into a functioning and evaluable AI application. Students compare predictive analytics, retrieval-augmented generation, agentic approaches, and workflow automation, then select suitable technologies and platforms for a defined use case. They specify data and interface requirements, use applied programming methods to construct a prototype, integrate it with a designated data source or system, and deploy it to a reproducible test environment. Baseline comparisons, task performance, operational value, reliability, safety, privacy, security, and relevant fairness considerations guide iterative improvement. Students examine constraints on scaling and prepare an operational handoff with monitoring and human oversight responsibilities. The course culminates in a working system, an evaluation report, and an individual defense of its business fit, technical choices, operating responsibilities, and deployment limitations.