I think you should only offer this service to people who are willing to first record the podcast in their own voice. Then your service could provide utility by replacing their voice to cleaning up their audio, removing speech impediments (e.g.
"uh uh um um um"), translating to other languages, editing in music, etc. Lots of people aren't good at reading/speaking so this service could be very useful to them, but you should make them prove the content is worth their own time by making them record at least a rough reading of the script themselves.
If the creator can't be bothered to read the script, it would be a waste of time for anybody else to listen to it.
