Auto-Learning shows you at a glance the pending work of your AI.
The Auto-Learning section now highlights what needs your attention for your AI to continue improving, without having to search through. On the dashboard’s front page, you will see a notice of “Pending AI Work“ (and an indicator in the side menu) that summarizes old unprotected errors, knowledge pending review, data conflicts, and suggestions awaiting approval. Within Auto-Learning, you will find new sections: a “Repair Queue“ that prioritizes what to reuse, modify, or clean; “Failed Regressions“ (past errors that are still not covered by an active rule); the rules that the AI retrieves but never ends up applying; and a summary of the “Learning Backlog.“ Additionally, suggestions that affect a workflow are clearly identified and lead you to review it instead of being applied blindly. These sections only inform you: they do not execute the AI or change your responses.