Thanks for having me. So the main thing from our side is that the latency budget for the dictation feature is about three hundred milliseconds end to end, and right now we are well over that.