I am very new to machine learning but I know that speech recognition is fairly easy to use with Tensorflow.js or ml5.js.
All of the examples I have seen use the micro