What you'll learn

Are you preparing for the Google Cloud Professional Machine Learning Engineer exam, or responsible for keeping production AI solutions safe, fair, and performant? In this course, (GCP-PMLE) Monitoring AI Solutions, you'll gain the ability to identify risks to AI systems and operationalize continuous monitoring on Vertex AI. First, you'll explore the threat landscape for AI workloads — from data poisoning and model extraction to prompt injection — and learn how to align your solutions with Google's Responsible AI practices, including assessing fairness, bias, and model explainability. Next, you'll discover how to configure Vertex AI Model Monitoring to detect training-serving skew and feature attribution drift, and how to establish continuous evaluation metrics that surface problems before they impact users. Finally, you'll learn how to monitor model performance against baselines and challenger models over time, and how to troubleshoot common training and serving errors using Cloud Logging and Cloud Monitoring. When you're finished with this course, you'll have the skills and knowledge of monitoring AI solutions needed to confidently operate production ML systems on Google Cloud — and to tackle this section of the PMLE exam.