MLOps Engineering on AWS
This 3-day course provides an overview of MLOps and the tools available to enable it in AWS.
The MLOps Engineering on AWS course is designed to enhance the DevOps approach in software development for building, training, and deploying machine learning (ML) models. It follows a four-level MLOps maturity framework, focusing on the first three levels: initial, repeatable, and reliable. The course emphasizes the critical roles of data, models, and code in successful ML deployments and highlights the importance of collaboration among data engineers, data scientists, software developers, and operations teams. Participants will learn how to use tools and processes to monitor model performance and address any deviations from key performance indicators.
This intermediate-level course lasts three days and includes various activities such as presentations, hands-on labs, demonstrations, knowledge checks, and workbook exercises. By the end of the course, attendees will be able to explain MLOps benefits, compare it with DevOps, assess security and governance needs for ML use cases, set up experimentation environments using Amazon SageMaker, and understand best practices for versioning and maintaining ML assets, including models and datasets. They will also learn about creating CI/CD pipelines in an ML context, implementing automated packaging and deployment, monitoring ML solutions, and automating the testing and deployment of models.
The course is aimed at MLOps engineers looking to deploy and monitor ML models in AWS and DevOps engineers responsible for maintaining these models. It is recommended that participants have completed AWS Technical Essentials, DevOps Engineering on AWS, or have equivalent experience, along with practical knowledge of Data Science with Amazon SageMaker.