Speech recognition software can analyze the sounds you make by filtering what you say, digitizing it to a format it can “read”, and then analyzing it for meaning.