Playbook
Our machine-learning team has outgrown notebooks and needs a proper platform
A retail bank's machine-learning team is running models in scattered Databricks workspaces with no central governance. They need a single managed catalogue, end-to-end model tracking, and a clear path to production generative AI on the same data.
Trigger — Production incidents in machine learning; the regulator wants lineage and access auditing.
Good outcome — Every model in the bank is registered in one place with a clear lineage back to the source data. The team can show the regulator who has access to what, and new generative-AI workloads run on the same governed platform.