Saving audio-records for later tests?

Question to the experts:

  1. Suppose: If Rhasspy is running in Docker, do I have a chance to adapt the code, so that the audio recordings are saved?
  2. If the recording is saved, e.g. as a wav-file, is it possible that Kaldi can interpret it?
  3. The result of this action is a json file, question: can this json also be saved so that it can be compared later?
  4. Is there anyone in the forum who has gained experience with it?
    Thanks in advance