Since 2016, Michelangelo has been the system that Uber engineers use to build, train, and deploy machine learning models at scale — from fraud detection to pricing to ETAs. Today we're making it available to everyone.

The open-source release includes the core platform: the pipeline framework (Uniflow), the API server, the controller, the local sandbox, and the CLI. You can run it on a laptop with ma sandbox create or on a Kubernetes cluster you already operate.

If you're interested in contributing, take a look at the contributing guide or browse the GitHub repository. For questions, open an issue or start a discussion on GitHub.