Advanced Engineering for MLOps on AWS
The course “DSAI307 – Advanced Engineering for MLOps on AWS” provides a comprehensive guide on implementing MLOps (Machine Learning Operations) using Amazon Web Services (AWS). Students will gain practical skills through a series of modules covering an introduction to MLOps, creating experimental environments with Amazon SageMaker, evaluating security and governance requirements for ML models, and best practices for versioning and maintenance models, implementing CI/CD pipelines, and monitoring ML-based solutions.
CODE: DSAI307
Category: Artificial Intelligence Course
DESCRIPTION
COURSE CONTENT
COURSE OBJECTIVES
ADDITIONAL INFORMATION
DESCRIPTION
Teaching methodology
The course includes educational laboratories in which each student will be able to carry out training exercises that will provide practical experience in the use of the tool, for each of the topics covered during the course.Prerequisites
- Basic knowledge of AWS.
- DevOps Engineering AWS.
- Completing DSAI107 course.
COURSE CONTENT
Below is an overview of the course contents:
- Introduction to MLOps: This module provides an overview of the fundamental concepts of MLOps (Machine Learning Operations), including the guiding principles, benefits and challenges. Students will gain an in-depth understanding of the role of MLOps in the machine learning lifecycle and large-scale model deployment.
- Experimental Environments in SageMaker Studio: Students will explore the experimental environment of SageMaker Studio, learning to create, configure, and manage development environments for machine learning. You will be provided with practical tools to experiment, iterate and optimize machine learning models in a controlled and scalable environment.
- Repositories: This module focuses on managing code repositories for machine learning projects. Students will learn best practices for organizing, versioning, and collaborating within repositories, using common tools and platforms such as Git and AWS CodeCommit.
- Orchestration: Students will gain skills in orchestrating and automating machine learning workflows using tools and frameworks such as Apache Airflow, AWS Step Functions, and Kubeflow. We’ll explore best practices for designing and running scalable and reliable machine learning pipelines.
- Scaling & Testing: This module focuses on scaling and testing machine learning models. Students will learn strategies for scaling models to handle large volumes of data and intensive workloads. They will also be introduced to model testing concepts, including unit testing, integration testing, and acceptance testing.
- Monitoring: Students will explore the importance of continuously monitoring machine learning models in production. They will be introduced to the concepts and technologies for monitoring performance, detecting errors, and maintaining model quality over time.
COURSE OBJECTIVES
At the end of the course, participants will be able to:
- Explain the benefits of MLOps.
- Compare and understand the differences between DevOps and MLOps.
- Create experimental environments for MLOps with Amazon SageMaker.
- Evaluate the security and governance requirements for an ML model and describe solutions and migration strategies.
- Explain best practices for versioning and maintaining model integrity.
- Describe three options for creating a CI/CD pipeline.
- Implement best practices to automate model deployment.
- Monitor ML-based solutions.
ADDITIONAL INFORMATION
Duration – 1 day
Delivery – in the classroom, on site, remotely
PC and SW requirements:
- Internet connection
- Web browser, Google Chrome
- Zoom
- Instructor: Italian
- Laboratories: English
- Slides: English