0 / 60 seg.

And then we used some algorithms that are a lot like speech recognition algorithms, called pattern recognition.