The ML Registry defines a centralized location for
- storing versioned machine-learning models and advanced analytics alongside parameters, metrics and other important artifacts
- data scientists to store information related to their machine-learning workflows
The ML Registry also provides a central component of any MLOps stacks built upon KX technology; a location to which information required for model monitoring can be stored, retrained pipelines can be pushed and models for deployment can be retrieved.
The functionality provides users of the KX Insights Platform and Microservices with
- A method of introducing their own models, generated prior to KX Insights, to the system with wrapped functionality to allow them to be integrated seamlessly with the requirements of the platform
- A way to understand the available models the registry knows
- A common storage location for all pipelines, q models, and Python models that can wrap the models so that on retrieval they are ready for deployment by KX Insights