Data Science and Machine Learning on Databricks

As the field of MLOps expands, data practitioners see the need for a unified, open machine learning platform where they can train, test and deploy models with as little friction as possible, along with tools like MLflow that simplify the ML lifecycle and make the process more rigorous and reproducible.

Using a real-world machine learning use case, you’ll see how MLflow simplifies and streamlines the end-to-end ML workflow. With MLflow on Databricks, you can use the MLflow Tracking server to automatically track and catalog each model training run through the data. This demo also shows how MLflow Projects neatly packages ML models and training environments into a universal project format, and how the MLflow Model Registry shepherds ML models through testing and staging environments into production, directly from within Databricks.