As of now, RC allows recorded phone calls to be downloaded in .mp3 format. Unfortunately, the compression this format sucks. RC needs to give serious consideration to providing alternative, lossless, formats such as flac. If you haven't heard, AI models for voice to text transcription have come a long way but using an mp3 file is terrible. Anyone looking to build a model would agree. If you don't catch up to this, I'm sure others will and a major reason why I would move to another platform.