One speaking step

In the app you answer with your own voice. Here you hear a sample learner, read by a synthetic voice. The feedback is what our app really wrote for that answer.

Your second try