Transcript cleanup has been a solved problem for a while. Visual cleanup is the actual bottleneck.
When you cut the audio, the video usually looks like a glitch in the matrix. Erasing the jump cuts entirely is a massive step forward for automated editing.
You nailed the point but stumbled on the words, so you record the whole thing again. And again
Speech Cleanup takes your first take and removes every filler, pause, false start, and retake automatically
The result is one seamless video. Other tools leave it jumping at the cuts