for that flat price, i bet they are already working on ffmpeg via webassembly. otherwise they will go under or dishonor those early clients pretty fast.

Aren't you just cutting the screen and playing a subway/gta/etc video below?

It can be done in any editor in a few clicks

Auto captioning will surely help, you can also provide styles for captions.

I remember revid.ai does similar stuff you can take inspiration from.

Download CapCut for free, drag both videos into your media panel, now click + on them both to drop them on the timeline. Make sure the ratio (under the preview window) is set to 9:16. Click on each video and move it around until you have it lined up where you want it. Export.

i was impressed with the features which kdenlive provides; had a small learning curve for some of the basics but went in expecting that