Why do knowledge responses differ between the Preview widget and other touchpoints?
In Genesys Cloud knowledge fabric, response variations can occur between the Preview widget and other touchpoints because they use different models for answer generation. Models are selected for each touchpoint based on specific use cases and requirements. For more information, see What types of knowledge connectors are available and when should they be used?
Some touchpoints also apply a confidence threshold in the background. If the system is not confident enough that an answer is accurate, the touchpoint does not return it. As a result, the Preview widget may show an answer while the touchpoint, such as an agentic virtual agent, returns no result.
A touchpoint can also modify a query before sending it to knowledge fabric, which may change the response.
To get the most accurate results, run the Preview tool using the touchpoint you plan to use in your organization. Then review your configuration settings and adjust them until the results meet your requirements.
[NEXT] Was this article helpful?
Get user feedback about articles.