Wondering if, you are planning to do something like OneTake AI do. breaking the screen to two and reposition (maybe with each paragraph?).
for example 16:9 horizontal left talking head and right the text on a brand coloured bakground (paragrah one), then swop them for the next paragraph. Same for Vetical top and bottom, then swop them after a given time. It would give a more dynamic interaction (the average attention span is 8.25 seconds), to the view to refocus again and again, rather than a TikTok looking running word that spread peoples attention between the listening and following up the spoken one word of the content.
2. Brand settings? ability to insert logo, change background colour and use of brand font (if not standard make font uploadable to HipClip)

3. I saw you planning the B-roll are there any more specific time (month or weeks) rather than "later this year"? I think this feature make or break your product moving forward fast.