No, we don't store answers, just the intent of the answer, which includes, 'Yes', 'No', 'Silence', 'Default'.
The slightly longer answer is that we take the audio, send it to a cloud service for Speech-to-Text Translation, then match the return of that (usually yes or no) to the intent. We store the text string of both the intent (yes, no, or unknown) and the string the speech service provided back ‘Yes I have’ or ‘No’. This is stored in our cloud service for the reporting feature.