Corey Zumar is a software engineer at Databricks, where he is working on machine learning infrastructure and APIs for managing the machine learning lifecycle with MLflow. He holds a master’s degree in computer science from UC Berkeley, where he was one of the lead developers of Clipper - an open-source project and research effort focused on high-performance model serving.
Simplifying Model Development and Management with MLflow
Introduced by Databricks in 2018, MLflow is the most widely used open source platform for managing the full ML lifecycle. With over 2 million PyPI downloads a month and over 200 contributors, the growing support from the developer community demonstrates the need for an open source approach to standardize tools, processes, and frameworks involved throughout the ML lifecycle. MLflow significantly simplifies the complex process of standardizing MLOps and productionizing ML models. In this talk, I’ll provide an overview of MLflow and discuss recent advancements in the platform, including simplified experiment tracking, new innovations to the model format to improve portability, new features to manage and compare model schemas, and new capabilities for deploying models faster.