Skip to Scheduled Dates
Course Overview
What if your machine learning models could move from development to deployment with the same confidence and consistency as your software releases?
MLOps Engineering on AWS is a hands-on, instructor-led course that helps you bring structure, automation, and scale to your ML workflows. Over three days, you’ll explore the use of tools and processes to monitor, deploy, and automate machine learning pipelines using AWS-native services like Amazon SageMaker, CodePipeline, and CloudWatch.
This course builds upon DevOps best practices and software development principles to help you move from experimental models to reliable production systems. You'll also discuss the use of tools and teamwork in addressing the challenges associated with ML deployment and monitoring.
Who Should Attend
This course is intended for any one of the following roles with responsibility for productionizing machine
learning models in the AWS Cloud:
DevOps engineers
ML engineers
Developers/operations with responsibility for operationalizing ML models
Course Objectives
By the end of the course, you’ll understand how to deploy machine learning models, take action when the model prediction in production drifts from agreed-upon key performance indicators, and automate the full ML lifecycle—from code to successful ML deployment in the AWS cloud.
- Implement DevOps best practices in machine learning workflows
- Design, deploy, and monitor secure, scalable ML pipelines
- Use Amazon SageMaker for experimentation, tuning, and deployment
- Automate CI/CD workflows for ML models, data, and code
- Take action when the model prediction drifts from KPIs
- Extend the DevOps approach to ML teams including data scientists, data engineers, and software developers
Course Outline
Day 1: Foundations of MLOps on AWS
- Overview of MLOps and its importance in deploying machine learning models
- Addressing the challenges associated with ML handoffs and teamwork
- Setting up secure environments and using Amazon SageMaker Studio
- Data versioning and pipeline structure
Day 2: Building and Automating ML Pipelines
- Designing CI/CD pipelines for ML deployment
- Automating model packaging, testing, and deployment
- Implementing security, A/B testing, and rollback strategies
- Real-world scenarios and best practices
Day 3: Monitoring and Operating ML Models
- Tools and processes to monitor and take action
- Model drift detection, alerting, and retraining
- Multi-account pipeline strategies
- Troubleshooting and human-in-the-loop feedback systems