Explore the current state of machine learning platforms and MLOps in this opening keynote from Databricks' co-founder and CTO Matei Zaharia and Director of Product Management Clemens Mewald. Delve into the complexities of operationalizing machine learning and discover common components of ML platforms. Learn about Databricks' approach to end-to-end data science and machine learning, including their unified data analytics platform powered by Apache Spark™. Gain insights into Databricks Notebooks, MLflow integration, Delta, Machine Learning Runtime, and model deployment. Understand how these tools and technologies work together to accelerate innovation by unifying data science, engineering, and business.
Overview
Syllabus
Intro
MLOPS VIRTUAL EVENT
A Common Story
Even After Deploying, Operating ML is Complex!
Why is ML Hard to Operationalize?
Response: ML Platforms
Common Components in an ML Platform
Our Approach at Databricks
End-to-End Data Science and ML on databricks
Databricks Notebooks Provide a collaborative environment for Unified Data Analytics
miflow Integration with Databricks Notebooks
A for Data Science and ML
mlflow Integration with Delta
Machine Learning Runtime
mlflow Integration with ML Runtime
mlflow Model Deployment
Taught by
Databricks
