Why can't we choose our own speech recognition engine in the official Pebble App?
I want to set custom timer names which are usually only one word by voice. Wispr Flow is just bad at recognizing the words (even if I set a fixed language), let alone the local engine. Why can't we just set our own speech recognition engine like microPebble allows us to do? Or just allow companion apps to use the voice stream and not just the transcribed text.