AI Development, Testing, Deployment

This unit guides students through hands-on practices for training, evaluating, and deploying AI products, emphasizing experimentation, bias mitigation, continuous data strategies, and trade-offs among performance, explainability, and fairness. It builds on foundational planning and technical concepts from earlier units and prepares learners to complete integrated prototypes and capstone deliverables in the final unit.

Learning Objectives

• Analyze trade-offs among model performance, explainability, and fairness and recommend strategies to balance these considerations for a given product requirement
• Apply model training, experimentation, and iterative prototyping techniques — including data collection and retraining plans — to develop and document an AI product iteration
• Demonstrate techniques to detect, evaluate, and mitigate bias in models using quantitative metrics and qualitative review, and propose remediation actions with measurable success criteria
• Evaluate deployment and collaboration workflows by designing an operational checklist that incorporates continuous data collection, automated drift detection/management, multimodal considerations, and team communication protocols