Through case studies and practical exercises, this seminar explores interface design patterns that makes complex machine learning models interpretable and interactive for general users. Discuss techniques for visually representing recommendations, inferences, uncertainties, and model insights to facilitate trust and adoption.