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?