Interesting angle. The tool I have doesn’t record or playback the audio. I’m using a transcribing engine to convert it to text to see what it captures. I figured that if a machine can figure it out, then almost any human can understand what I said.

I then use a summarizing machine learning algorithm to see if the output matches the summary of what I wanted to convey.