We’ve added a new view that allows you to group experiment results by metadata. Compare the performance of your evaluations across different segments (eg. user segments or subject areas) to pinpoint what’s going well and where your application can improve.

Check out our documentation for more information.