It takes a video file, converts it into an audio file, transcribes the audio to text using OpenAI's Whisper API and identifies segments between specified code-words. It then cuts out these segments from the original video, removes any silence and produces a new edited video file ready for the final touches. False recording starts are also okay as it only matches the last occurrence of the 'start' code-word when the next code-word is the 'end' code-word.

I wanted to jump into the far deepest of deep ends and ask for feedback please.

I am kind of scared of you all tbh but I value learning more than crapping my pants.

Thank you.