I noticed a recent change in how Alexa processes user requests that is causing negative behavior in my skill. My application has both 'one-shot' and 'interactive' modes of user interaction. When a user is in interactive mode, Alexa has no problem understanding requests. However, when a user makes a request in one-shot mode, Alexa fails to parse the utterance correctly more than half of the time, which results in my application failing to return the desired information. What can I do to improve this?
message:
I noticed a recent change in how Alexa processes user requests that is causing negative behavior in my skill. My application has both 'one-shot' and 'interactive' modes of user interaction. When a user is in interactive mode, Alexa has no problem understanding requests. However, when a user makes a request in one-shot mode, Alexa fails to parse the utterance correctly more than half of the time, which results in my application failing to return the desired information. What can I do to improve this?