We prototyped a voice bot flow for EN-US. This works, and all is well in our testing. However, when we go to add other English dialects to the flow, we have to start from scratch on pre-adding utterances. Is there no way to copy utterances across dialects of the same base language?
This raises the bar pretty high to properly support multiple dialects of the same language because it requires utterances to be hand-keyed per-dialect, or for intent miner to be run for each dialect and the NLU be trained per dialect. I can understand not supporting this to copy utterances between different base langauges (ex., English to French), but not between dialects (EN-US to EN-AU).
#ConversationalAI(Bots,AgentAssist,etc.)#Implementation------------------------------
Paul McGurn
Manager, Telecom & DevOps
GoTo
------------------------------