It might be nice to create an audio node with the pipeline as well as a TTS tool like [Deepgram](https://developers.deepgram.com/docs) into the pipeline. Here's an example pipeline 