Model Prototyping and Deployment

This unit covers rapid prototyping, visualization, deployment, and monitoring of machine learning models. Students will build interactive demos, create publication-quality graphics, deploy models as APIs (including a simple random forest), and implement basic monitoring strategies for LLMs and traditional models to prepare for production work in the final unit.

Learning Objectives

• Apply rapid prototyping techniques using Gradio to build interactive model demos and collect user feedback
• Create informative visualizations with Plotnine to communicate model performance, feature relationships, and data distributions
• Demonstrate deployment of machine learning models as RESTful APIs, including packaging a simple random forest model and exposing endpoints
• Evaluate model behavior in production by implementing monitoring for LLMs and traditional models, defining key metrics, detecting drift, and specifying alert/response strategies