Vertex AI - Pipelines for Your MLOps Workflows

Explore Vertex AI Pipelines for MLOps workflows in this 31-minute conference talk from Devoxx. Learn how this managed unified ML platform streamlines machine learning processes for data scientists and engineers. Discover the benefits of experiment pipelining beyond traditional model development stages. Examine practical examples showcasing Vertex AI Pipelines' developer-friendly features and its ability to meet custom ML needs. Gain insights into the platform's comprehensive toolset for machine learning workflows, including data preparation, model training, hyperparameter tuning, validation, and deployment. Understand how Vertex AI's managed serverless approach eliminates infrastructure management overhead, making it ideal for teams without dedicated DevOps or sysadmin engineers. Explore topics such as ML Ops, conditional steps, triggers, and the advantages of Vertex AI's fully managed resources, including an ML metadata store and feature store.

Intro
ML Ops
What is Vertex AI
Why are pipelines useful
Example
Conditional steps
Triggers
Conclusions
Outro