Feedbacks
The Feedbacks tab shows every individual rated conversation turn (thumbs-up or thumbs-down) together with the full question and answer text. It is available on both standard and orchestrator domains. The table columns are:
- Question — the user's question.
- Normalised Question — the canonical form of the question.
- Answer — the answer the chatbot returned.
- Liked/Disliked — the rating, shown as a color-coded icon.
- Comment — optional free-text the user left with the rating.
- Updated At — when the rating was last changed.
The header row provides filters for creator, question text, answer text, feedback comment, liked/disliked status, and date range. Rows load via infinite scroll as you scroll down, and clicking any row opens a detail drawer showing the full turn.
Tip
Feedbacks vs. Human feedback: the Human feedback tab shows
aggregated per-document scores and rankings, while Feedbacks shows the raw individual rated
turns — use it for deeper, turn-by-turn investigation.
FAQ
What does the Feedbacks tab show?
The Feedbacks tab shows every individual rated conversation turn (thumbs-up or thumbs-down) along with the full question and answer text. It is available on both standard and orchestrator domains. It’s intended for investigating feedback at the level of individual turns.
What columns are included in the Feedbacks table?
The table includes Question, Normalised Question, Answer, Liked/Disliked, Comment, and Updated At. Question is the user’s question, Normalised Question is the canonical form, and Answer is what the chatbot returned. Liked/Disliked is shown as a color-coded icon, Comment is optional free text, and Updated At shows when the rating was last changed.
How do I filter and browse rows in the Feedbacks table?
Use the header row filters for creator, question text, answer text, feedback comment, liked/disliked status, and date range. Rows load via infinite scroll as you scroll down. Clicking any row opens a detail drawer showing the full turn.
What is the difference between Feedbacks and Human feedback?
Human feedback shows aggregated per-document scores and rankings. Feedbacks shows the raw, individual rated conversation turns. Use Feedbacks when you need deeper, turn-by-turn investigation.