DevOps for Machine Learning
summary
Shows a concrete DevOps-style ML pipeline: code and ETL in git, infrastructure-as-code to spin GPU VMs, training jobs that register model artifacts, and automated release to Docker/Kubernetes. Cites DORA metrics (high performers deploy 46× more frequently, have ~2,500× faster lead time and 7× lower failure rates). Covers model versioning, artifact storage, commit- or data-triggered builds, monitoring for data drift, and automated continuous retraining with cost-aware short-run tests before full training.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:30Machine Learning
- 5:40Artificial Intelligence vs Machine Learning
- 11:24Why DevOps
- 12:47Innovation and Stability
- 14:04DevOps Definition
- 19:48Experimentation
- 20:46Software Development
- 21:41Planning
- 27:51Source Control
- 31:10Continuous Integration
- 37:33Monitoring and Learning
- 40:58Pipeline
- 45:53Examples
- 54:03Summary
related talks
57mDevOps for Machine Learning
59mDevOps for Machine Learning Projects
01mMLOps: Applying DevOps to ML
44mDevOps for Machine Learning - why is it different?
29mDevOps for Data Science? - automate the boring stuff and leverage the OSS ecosystem
21mMastering the DevOps Machine Learning Pipeline
26mApplying DevOps practices in Data and ML Engineering
39mContinuous Delivery for Machine Learning — The Future of MLOps