Personalizing recommendations with feedback

• Default on 

Who: Employee

Requirements: Admins must define specific actions for courses, jobs,  projects  or to ensure the feedback options appear correctly on recommendation cards.

What's new: We have introduced a robust feedback loop built on a new, scalable data architecture that allows users to explicitly influence their recommendation feeds. This update provides greater user control over AI discovery by providing a mechanism to hide irrelevant content. 

How it works: Users can select "Interested" or "Not Interested" from the overflow menu of recommended items to trigger a feedback modal. Selecting "Not Interested" and checking "Do not show recommendation" removes the item from the feed upon submission; otherwise, no changes occur. Responses can be updated by switching options or removed by deselecting the current choice, instantly updating the feed and recommendation signals. While users can deselect previous choices, dismissed items may become hidden and uneditable, as there is currently no centralized view for all preferences. To understand how this works, read Manage recommendation feedback.

Value: AI-based Recommendations