Possible to make audio from .srt file with sample audio from timestamp?

[ageekhere]

This is a bit of a long shot but here we go.

The scenario
Have a show that is in another language with subtitles and a .srt file with time stamps.

The goal
To have the ai convert the sub titles into AI voice while keeping the audio in sync with the subs and video like https://speechgen.io/en/subs does it.

But instead of using generic AI voices have the AI use the audio sample at the timestamps in the srt file so the AI voice which will be in English could sound like the person in the show which is speaking a different language.

Following the effort here https://www.reddit.com/r/LegoMasters/comments/18jz5vk/lego_masters_subtitle_project/
The idea is to create an English dub using AI voices to sound like the people in the show with the srt files and time stamps.