feathr-ai/feathr

This system is a feature store platform that enables users to define, manage, and serve machine learning features across diverse data sources and execution environments. It provides a Python SDK and CLI for feature engineering, with support for offline and online stores, Spark-based processing, and integration with Azure Purview and SQL-based registries. The platform includes a web-based UI for visualization and management, along with role-based access control and automated testing infrastructure.

The project underwent a major Python SDK refactoring to introduce a Pythonic API and a new compute graph model, while simultaneously expanding the feature registry with SQL, Purview, and RBAC-backed implementations. This period also saw the migration from SBT to Gradle, the addition of a web UI, and the removal of legacy Scala components and build files.

48 changes

This period focused on improving the developer experience by introducing a local Feathr sandbox environment and expanding the Python client with a DSL generation API. The work also included adding test coverage for registries and integrating an experimental chatbot for notebook interactions.

4 changes

feathr-ai/feathr was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point the surveyor at a repository you know and see whether you agree with it.