DevOps for Data Science With Kubernetes
summary
Demonstrates a pipeline that takes Git-hosted Jupyter notebooks, uses a source-to-image builder to extract dependencies, run training and produce a containerized model service endpoint deployable on OpenShift/Kubernetes, successfully queried from another notebook during the demo. Shows MyBinder for reproducible notebooks, Seldon for serving, Prometheus for visualizing prediction distribution metrics to detect data drift, and claims Kubernetes gives self-service GPU access, portable reproducible environments, and easy incremental rollouts of new models.
AI-written summary · about
chapters (5)
- 0:00Start
- 9:21Immutable images
- 10:05Stateless microservices
- 12:17Declarative app configuration
- 17:25More flexible: source-to-image
related talks
36mData Science Meets DevOps: Gitops with OpenShift
30mContainer workflows for data science and ML
57mDevOps for Machine Learning
33mCompliant Data Management and Machine Learning on Kubernetes
17mProductionizing Data Science Workloads
59mDevOps for Data Science
35mCase Study: AI-as-a-Service on Kubernetes at Scale and In Production
06mEvolving with Kubernetes: Embracing Model Ops