Does it depend on the source material though? I think I'm happy with scattered tone and pitch if it comes out funny

Take this video of my now ex prime minister: https://www.youtube.com/watch?v=aHsFtANY5Ro (little bit of NSFW language)

Yes it definitely depends on the source material… I’ll see about adding it in.

Amazing! Love it so much. My brain was running wild with possibilities (like having an autocomplete from the corpus, live audio only previews, and the above)

I didn't realise there was a github! You should add a link to your tutorial

https://github.com/antiboredom/videogrep

If I have the time I'll see how I go dipping my toes in the code.

I guess there would need to be an intermediate step. Videogrep helps to surface useful ngrams and there would still be a manual/creative step to stitch them together in a way that works.