Has anyone had success with keeping barge-in enabled, and accurately having the bot detect a barged-in utterance? My testing is reliably failing in that the first word or two get chopped off, so the utterance won't match the slot. Disabling barge-in leads to annoyance because repeat callers expect to be able to utter the same thing they did previously, but they're forced to wait for the prompt to finish. Leaving it on leads to No Match paths being hit, instead.
#ArchitectureandDesign#ConversationalAI(Bots,AgentAssist,etc.)------------------------------
Paul McGurn
Manager, Telecom & DevOps
GoTo
------------------------------