Audio

Either:

• HumanProvidedAudio->

  • AiTranscriptionText (via Whisper )->

    • LLMChain->

      • AiResponse, or

• HumanProvidedText->

  • AiGeneratedAudio (e.g. via OpenAi)