Configure Parameters
JSON type object
configureParams JSON type objectUse configureParams to set Conversation behavior that applies to later operations. Individual modules can override settings where supported.
Behavior and notes
- Use conversational audio processing for operations that intentionally recognize or record while prompt audio is playing.
Developer guidance
Configure Parameters sets values used by later operations. Runtime fields are named values that can be referenced by supported properties using $[FieldName]. The runtimeFields array accepts string, integer, boolean, and floating-point values. A nested object such as speechSynthesisVoice is configured through its own properties; it is not replaced as one scalar runtime field.
Example in context
Illustrative JSON excerpt (not a complete configuration document):
{
"configureParams": {
"runtimeFields": [
{
"name": "CustomerName",
"value": "Antoine"
},
{
"name": "RetryCount",
"value": 3
},
{
"name": "IsConfirmed",
"value": true
},
{
"name": "Confidence",
"value": 0.95
}
],
"speechRecognitionLocale": "en-US",
"speechSynthesisVoice": {
"name": "Ava"
}
}
}
This excerpt assigns four typed runtime fields and configures recognition locale and a named voice. Only individual supported scalar properties accept $[FieldName] substitution.
Properties
Jump to a property:
resourceLanguage · runtimeFields · speechSynthesisEngine · speechSynthesisVoice · speechRecognitionEngine · speechRecognitionLocale · recognitionContext · speechSynthesisAliases · speechSynthesisPhonemes · playThenRecordEnabled · playThenRecognizeEnabled · audioOutput · audioProcessingModeresourceLanguage
Sets the resource language used to locate Conversation resources.
runtimeFields
Sets or updates Conversation runtime fields. Values persist and may be referenced with $[FieldName]. Each array element is a Runtime Field Entry containing a name and a value. Supported values include strings, integers, booleans, and floating-point numbers.
speechSynthesisEngine
Sets the default speech-synthesis engine for subsequent operations.
speechSynthesisVoice
Sets the default speech-synthesis voice.
speechRecognitionEngine
Sets the default speech-recognition engine for subsequent operations.
speechRecognitionLocale
Sets the default speech-recognition locale, for example en-US.
recognitionContext
Defines recognition-context terms applied to recognition tasks.
speechSynthesisAliases
Defines speech-synthesis text aliases.
speechSynthesisPhonemes
Defines speech-synthesis phonetic mappings.
playThenRecordEnabled
When true, playback completes before recording begins when the operation would otherwise overlap.
playThenRecognizeEnabled
When true, playback completes before recognition begins when the operation would otherwise overlap.
audioOutput
Selects the audio output behavior.
systemDefault— Uses normal system audio routing.deviceSpeaker— Uses the device's built-in speaker; not applicable on macOS.
audioProcessingMode
Selects audio processing for capture and playback.
standard— Standard audio processing.conversational— Enables voice-processing behavior suited to overlapping playback and capture, including echo cancellation.