Allowing sound to pass on from external microphone connected to Raspberry PI and into a trained neural network.

I have a ICS‐43434 mems microphone connect to my Raspberry Pi GPIO pins.

I also have a trained classification model.

I was planning to deploy the classification model as a standalone executable on the raspberry pi. The Pi will then allow sound to pass through the external microphone and into the classification model for classification purpose.

I have read and understand how to deploy the model as a function into the pi. But i'm not too sure about how to read the audio digital signal from the mems microphone and do feature extraction on it to feed into the model.

Appreciate any help if someone could kindly guide me through the process.