interface AgentTranscriptionOptions {
    agentTranscription: boolean;
    agentTranscriptionSpeech: number;
    hyphenateWord: ((word) => string[]);
    sentenceTokenizer: tokenize.SentenceTokenizer;
    userTranscription: boolean;
    wordTokenizer: tokenize.WordTokenizer;
}

Properties

agentTranscription: boolean

Whether to forward the agent transcription to the client

agentTranscriptionSpeech: number

The speed at which the agent's speech transcription is forwarded to the client. We try to mimic the agent's speech speed by adjusting the transcription speed.

hyphenateWord: ((word) => string[])

A function that takes a string (word) as input and returns a list of strings, representing the hyphenated parts of the word.

Type declaration

    • (word): string[]
    • Parameters

      • word: string

      Returns string[]

sentenceTokenizer: tokenize.SentenceTokenizer

The tokenizer used to split the speech into sentences. This is used to decide when to mark a transcript as final for the agent transcription.

userTranscription: boolean

Whether to forward the user transcription to the client

wordTokenizer: tokenize.WordTokenizer

The tokenizer used to split the speech into words. This is used to simulate the "interim results" of the agent transcription.