Implement MLOps practices with Amazon SageMaker
summary
Shows how SageMaker Pipelines and a model registry automate end-to-end MLOps: project templates spin up repos and CI/CD, pipelines run preprocessing, distributed training, evaluation and conditional model registration, and CodeBuild/CodePipeline handle promotion to staging/prod across accounts. Demo runs a ~10-minute pipeline, logs RMSE 2.3 on validation, lets reviewers approve versions, and retrains on a git commit or data-trigger while preserving full lineage and artifacts, plus the ability to stop, debug and resume steps.
AI-written summary · about
chapters (7)
- 0:00Intro
- 2:40Overview
- 10:10SageMaker Pipelines overview
- 15:24Troubleshooting
- 18:30MLOps Templates
- 21:00Demo
- 32:10Resources
related talks
34mImplement MLOps practices with Amazon SageMaker
29mImplementing MLOps practices with Amazon SageMaker
36mImplementing MLOps practices on AWS
29mHow to create fully automated ML workflows with Amazon SageMaker Pipelines
37mStreamlining Machine Learning Operations
33mBuilding & Managing MLOps Pipelines in the Cloud
45mAutomate your ML workflows with pipelines
37mStreamlining Machine Learning Operations