Speech recognition flow nodes

Configure recognition and start listening before running these triggers. See how nodes work for the diagrams.

Sentence Recognized

Fires for each sentence recognized.

Sentence RecognizedTriggeredTextConfidenceTimestamp (UTC)CultureWord CountContext
Outputs
TriggeredFlow
TextString
ConfidenceNumber
From 0 to 1.
Timestamp (UTC)String
CultureString
The language, such as en.
Word CountNumber
ContextString
Details about the model and audio used.
Properties
Min Confidence
Ignore sentences below this confidence, 0 to 1. Default 0.
Text Filter
The text to match.
Match Mode
Any, Contains, or Exact. Default Any, which ignores Text Filter.
Case Sensitive
Default off.

Exact compares the filter with the whole sentence.

Word Recognized

Fires once for each word in a recognized sentence.

Word RecognizedTriggeredWordNormalized WordWord IndexSentenceConfidenceTimestamp (UTC)CultureContext
Outputs
TriggeredFlow
WordString
Normalized WordString
The word in lowercase.
Word IndexNumber
Position in the sentence, counting from 0.
SentenceString
ConfidenceNumber
The sentence's confidence.
Timestamp (UTC)String
CultureString
ContextString
Properties
Min Confidence
Ignore words below this confidence, 0 to 1. Default 0.
Word Filter
The word to match.
Match Mode
Any, Contains, or Exact. Default Any, which ignores Word Filter.
Case Sensitive
Default off.

To react to a keyword, set Word Filter to the word and Match Mode to Exact.

Contributors

Thank you to everyone who has helped improve this documentation.

bwmpbwmp