event · community record
MODEL VERSIONING: WHY, WHEN, AND HOW
MODEL VERSIONING: WHY, WHEN, AND HOW was a developer community event in San Francisco.
01
Talks & recordings
2 connected sessions
Model Monitoring in Production
Real mission critical AI systems require advanced monitoring and model observability that enables continuous and reliable delivery of ML models into production. This demo-based talk covers data drift detection, model degradation, and the full cycle of ML in production with Kubeflow pipelines.
Model Versioning: Why, When, and How
Models are the new code. In the wild-west of data science and ML tools, versioning, management, and deployment of models are massive hurdles. Drawing on experience with ModelDB and Verta, this talk presents best practices and tools for model versioning—a Git for ML models—enabling rapid deployment and high quality of deployed ML models.