I have an awesome idea. You want to give podcast experience. real podcast experience is achieved through person's voice. so a simple tool to train custom voice model of user or just switch AI voice model might be helpful.
Thank you for sharing the background working of hashnode.