AWS ML BlogSaturday · May 30, 2026FREE

Streamline external access to Amazon SageMaker MLflow using a REST API proxy

The AWS Machine Learning Blog published a post on May 28, 2026, detailing how to build a secure Flask-based MLflow proxy service for Amazon SageMaker. The proxy provides HTTPS access to SageMaker MLflow without requiring the MLflow SDK, allowing external tools and scripts to interact with MLflow via REST API calls. This is particularly useful for organizations undergoing cloud transformation who want to preserve existing ML workflows while adopting cloud-native services. The solution involves creating a Flask application that authenticates requests using AWS Signature Version 4 and forwards them to the SageMaker MLflow endpoint. The post includes step-by-step instructions, code snippets, and considerations for security and scalability. By using this proxy, teams can avoid modifying their existing codebases to use the MLflow SDK, simplifying the migration process. The proxy can be deployed on AWS services like Amazon ECS or AWS Lambda, and it supports custom domain names and SSL certificates for secure access.