Amazon.com, Inc.
Speech-processing system

Abstract:

A system may receive audio data that represents a wakeword associated with a first speech-processing system and a command associated with a second speech-processing system. Different indications of handing the audio data off to the second speech-processing system may be determined based on a determined amount of interaction with the second speech-processing system. If the amount of interaction is low, a longer, more detailed indication is generated; if the amount of interaction is high, a brief, less detailed indication is generated. A local device may output audio corresponding to the indication before outputting audio generated by the second speech-processing system in response to the command.