Vatsal-D07/Yapp.ai ? reverse-engineered prompt

Build me a web app that turns a written script into a podcast or short reel voiceover.

I want to paste in text, choose a voice, and generate an audio file I can play right away, preview with a waveform, and save to my account so I can come back to old jobs later. It should support multiple languages, feel polished, and include a few preset voices plus voice cloning if that’s available. Make it work nicely in the browser, with a clean simple UI, and show progress while the audio is being created.

It would be great if the app works in a fully offline demo mode too, so I can test everything without setting up any accounts. If you need to look up current docs online for anything, go ahead and do that. Please wire up the backend, storage, and basic job status handling so the whole flow works end to end.