Can I trust the accuracy of AI CoPilot? Is there a way to flag when an incorrect response is provided?

While the Heap team has made a significant effort toward creating a consistently accurate chat interface, the underlying AI technology (large language models) is occasionally susceptible to “hallucination”, which may result in incorrect responses. We recommend users double-check any results provided by AI CoPilot.

To provide feedback, ‌click the “thumbs up/down” buttons and please share any context you can on the desired functionality. Our team regularly checks this feedback and uses it to improve the quality of the AI CoPilot experience.

AI CoPilot thumbs up/down response feedback
Last updated