What is the point of the model staging and promotion functions in MLflow?

User16826992666
Databricks Employee
Databricks Employee

Why not just directly deploy the model where you need it in production?

sean_owen
Databricks Employee
Databricks Employee

The Model Registry is mostly a workflow tool. It helps 'gate' the process, so that (for example) only authorized users can set a model to be the newest Production version of a model - that's not something just anyone should be able to do!

The Registry can trigger webhooks when a model enters a new state, for example to run a testing notebook when a new Staging candidate appears.

It also helps serve as a single source of truth: what's the current production model? rather than pass around the run ID "abc123456..." informally, the stage tags in the Registry let any process look up the 'latest production model' reliably.

View solution in original post