Leveraging AWS and Kubernetes for Scalable AI/ML Workloads in the Cloud
summary
Shows a working workflow that deploys infra with CloudFormation, uploads data to S3 which triggers a Lambda to start a SageMaker pipeline that writes features into a Feature Store, and runs training/serving on EKS via EKS + Kubeflow/KFServing; CI/CD uses CodeCommit, CodeBuild and CodePipeline; monitoring relies on CloudWatch; recommends Spot Instances for batch jobs to cut costs. Demonstrates end-to-end automation and where bottlenecks and cost savings actually appear rather than abstract benefits.
AI-written summary · about
chapters (10)
- 0:00AWS & Kubernetes intro for AI/ML
- 2:01Automated AI/ML workflow with AWS
- 3:05Benefits of AWS for AI/ML
- 4:34Kubernetes advantages for AI/ML
- 6:01AWS + Kubernetes for scalability
- 7:29Best practices for AI/ML workloads
- 8:57Real-world AI/ML use cases
- 10:02Kubeflow with Amazon EKS
- 11:27Summary and key takeaways
- 12:12Q&A
related talks
25mDesigning an AWS Well Architected Framework for AI/ML
34mBuilding End-to-End Machine Learning Workflows with Kubernetes...
40mBuilding scalable end-to-end deep learning pipelines in the cloud
34mDeploying a Machine Learning System That Only Runs Locally
34mBuilding scalable end-to-end deep learning pipelines in the cloud
43mMachine Learning Ops: DevOps for Data Science
42mKubernetes for Machine Learning: Managing and Scaling AI Workloads
54mScaling Machine Learning Pipelines in Cloud