Welcome to the AWS SageMaker course, where you'll learn utliize AWS's machine learning capabilities for building and hosting models efficiently. This course is designed for both AWS novices and AWS seasoned professionals, who are starting out on their journey to developing the ML skills to manage, deploy, and scale machine learning projects using SageMaker's features. Even if you have no ML or SageMaker experience, the course will jump start your ML learning. The course will follow a typical machine learning pipeline from data preparation all the way through to hosting and monitoring. Activities in the pipeline align to different personas and so the course follows the persona actions for each stage.
- Pre-requisites: This section outlines the essential foundation needed for your SageMaker learning journey. You'll review machine learning basics, understand the necessary mathematical concepts, and learn why SageMaker may initially seem complex. The section also emphasizes the advantages of learning through persona-based actions, ensuring a targeted and effective educational experience.
- SageMaker Introduction: Dive into SageMaker as a powerful managed service. This section introduces key personas—data engineers, scientists, and MLOps engineers—showcasing SageMaker’s versatility. Learn about Jupyter Notebooks, working locally first and then remotely hosted, and the benefits of the SageMaker SDK for Python over other tools. You'll also explore data preparation essentials for ensuring you have high-quality data that is ready for model training.
- SageMaker UI: Mastering the SageMaker user interface is crucial. This section guides you through UI navigation, comparing legacy notebooks with SageMaker Studio, and exploring code editor alternatives. Understand the differences between SageMaker Studio Classic and the new version, enabling you to optimize your workflow and collaboration efforts.
- Persona SageMaker Activities - Data Engineer: This section equips data engineers with tools for large-scale data preparation and management. Explore tabular data preparation, SageMaker Canvas, AutoML, and Jupyter Notebooks for data processing. Gain skills to streamline data workflows, ensuring efficient transformation of raw data into actionable insights.
- Persona SageMaker Activities - Data Scientist: Tailored for data scientists, this section covers feature engineering, model training, and optimization in SageMaker Studio. Learn to manage experiments and track models using the SageMaker Model Registry. Practical activities ensure you're ready to enhance model accuracy and streamline deployment.
Persona SageMaker Activities - MLOps Engineer: MLOps engineers will learn to manage and deploy models effectively. Explore hosting options, advanced inference, and automating pipelines with SageMaker. Integration with tools like Apache Airflow and model monitoring strategies prepare you to manage scalable, reliable ML workflows in production.
