Currently, the Sentiment Agent in Thread processes sentiment at the time of each interaction, but it doesn’t automatically re-run or update the score after final notes or additional feedback are added later. This means that if positive resolution notes come in after an initial negative interaction, the sentiment score may not reflect the full story, and there’s no built-in mechanism to trigger a re-evaluation.

Even Ask Magic can see the final notes and recommend a sentiment revision, but the core sentiment score doesn’t update, and highlights a real gap.

Requesting Both:

The ability to manually trigger a sentiment re-evaluation on a ticket, and/or

Automatic reprocessing of sentiment when final notes or CSAT feedback are added?

Basically a Sentiment Recap.

Bad data isn't good for any of us: faulty data or odd interpretation can't ever be corrected currently. The bad sentiment information is hard coded and will affect overall company averages forever.

Please fix.