Here I\'m trying to replicate what I would be doing on an ESP32 chip for streaming audio to wit.ai\'s speech recognition API (this is why I\'m not using some higher level module