Model Deployment and Monitoring

This unit focuses on turning trained models into usable services and interfaces, covering rapid prototyping, visualization, deployment, and monitoring practices. Students will build interactive demos, produce communicative graphics, deploy models as APIs, and establish basic monitoring for both traditional models and LLMs.

Learning Objectives

• Apply rapid prototyping techniques using Gradio to build interactive ML demos that integrate model inputs, controls, and outputs
• Create reproducible visualizations with Plotnine to communicate model predictions, evaluation metrics, and data distributions to stakeholders
• Deploy machine learning models (including a Random Forest) as RESTful APIs, document endpoints, and perform functional testing of requests and responses
• Implement and evaluate basic model and LLM monitoring workflows by instrumenting logs, defining performance thresholds, and identifying drift or anomalies