End-to-End MLOps with MLflow and Kubeflow - Nick Chase, CloudGeometry

End-to-End MLOps with MLflow and Kubeflow - Nick Chase, CloudGeometry
Machine Learning Operations (MLOps) ensures the smooth transition of ML models from development to production. With so many tools available, how do you choose the right one and integrate them effectively? Enter MLflow and Kubeflow - two powerful tools that complement each other in the MLOps landscape. This talk provides a hands-on guide to implementing an end-to-end MLOps pipeline using MLflow for experiment tracking and Kubeflow for scalable deployment and orchestration. We will cover:
Introduction to MLflow and its components.
Experiment tracking, model versioning, and reproducibility with MLflow.
Introduction to Kubeflow's ecosystem.
Model deployment, scaling, and orchestration using Kubeflow.
How MLflow and Kubeflow complement each other.
Setting up an end-to-end pipeline: from experiment tracking with MLflow to deployment with Kubeflow.
Live demo
You will leave this talk ready to set up your own MLOps pipeline with MLflow, Kubeflow, or both, as appropriate.