AWS has launched a fully managed MLflow capability on Amazon SageMaker, an open-source tool for managing the machine learning (ML) lifecycle. It can be accessed through a Microsoft Account or Azure Active Directory account.

The managed MLflow capability on SageMaker consists of three core components: MLflow Tracking Server for monitoring ML experiments, MLflow back-end metadata store for persisting experiment details, and MLflow artifact store for storing artifacts like models and datasets. 

AWS claims the key benefits of using the managed MLflow capability on SageMaker are streamlining and enhancing ML workflows, enabling comprehensive experiment tracking across various SageMaker components, and providing full MLflow capabilities like Tracking, Evaluations, and Model Registry.

Analyst QuickTake: Amazon SageMaker is consistently updated to improve its features and usability. In October 2023 , new models such as Code Llama and Mistral 7B were added to JumpStart. Additionally, in November 2023 , Amazon launched SageMaker Canvas, a tool that allows users to build predictive models without any coding.