AI Sync Picture

I did a live stream that we also recorded to disk at 4k MOS.  Now in post we need to sync the 4k picture to the 1080 stream, but we don't have an audio reference.  It would be pretty sweet if we could just Sync Picture.  With all the generative video and facial recognition stuff out there, it seems like a no-brainer in terms of technical hurdles.  It's almost like a proxy workflow except it's two separate files that need to be matched.