Product capability
Response Evaluation
Response Evaluation checks an AI answer against defined quality KPIs and summarizes the result in one overall score.
The feature works as a second review step for AI output. It highlights whether the answer is strong enough to process further, where the weak points are, and what the user should review next.
Overall score
Summarize answer quality in one score
The overall score summarizes the evaluation across the active KPI set. It gives users a quick signal for how reliable, complete, source-aware, and instruction-fit an answer is before they reuse it.
- Review the active KPI set
- See one overall quality signal
- Check reliability before reuse
Detail report
Open the evaluation as a separate report
The detail report opens as a separate view. It shows KPI values, source signals, the evaluated message, metadata, and model-specific results so the user can inspect why the score was assigned.
- Inspect KPI and source signals
- Retain the evaluated message
- Understand model-specific results
Wrap-up
Show the short summary and next steps
The compact wrap-up gives a short summary and recommended next steps directly near the evaluated answer. It turns the review into practical guidance instead of leaving the user with only a score.
- Read a compact summary
- Act on recommended next steps
- Move beyond an isolated score