NLP Confidence Threshold Override in Architect Flows

  • Encountering inconsistent intent matching when setting confidence thresholds within the Genesys Cloud Architect flow.
  • The bot falls back to the default path despite the NLP engine returning a score of 0.85, which exceeds the configured 0.80 limit.
  • No error logs appear in the interaction detail view, yet the business logic branch is not triggered as expected.