Operational AI Development and Evaluation

This unit teaches students how to operationalize AI products by translating research and model design into repeatable training, experimentation, and deployment practices. Students will integrate technical workflows with cross-functional collaboration and business insight to prepare models for continuous improvement, fairness, and real-world use ahead of final product maturation.

Learning Objectives

• Analyze trade-offs between model performance, explainability, fairness, and business objectives to select appropriate architectures and implementation strategies
• Apply the end-to-end model training lifecycle—data preparation, experiment setup, iterative training, validation, and automated retraining pipelines—to manage model drift and sustain performance
• Demonstrate the design and execution of integrated experimentation (A/B tests, ablation studies) and prototype iterations, documenting metrics and collaborating across technical and business teams to inform roadmap decisions
• Evaluate ethical, bias-mitigation, and multimodal design decisions (including generative vs. traditional approaches) and produce actionable business insights to communicate outcomes to stakeholders