The TTS is pretty solid, what did you use for that? Amazon Polly is the best options I know of personally, but I'd love to know what you use because it's pretty solid.

Thanks! Actually I keep worrying that the TTS is the weakest link. Good to know that there are people who feel it's up to the mark.

I'm using Google Cloud's Text-To-Speech - one of the WaveNet models over there. Haven't tried Polly at all! Perhaps I ought to explore that a bit.

I love this idea. How hard would it be to adapt it to eprint.iacr.org?

Thanks! Are the papers on iacr.org different from what gets uploaded to Arxiv?

I need paper metadata to be available in a structured format. I currently support the "arXivRaw" metadata format of the OAI-PMH protocol [1]. Unfortunately I didn't support RSS yet. Are you one of the maintainers of iacr? What is the size of its userbase?

[1] - https://arxiv.org/help/oa

iacr.org is for cryptography-related papers, and a lot of pre-prints go there. I think a lot of them don't get to Arxiv.

Not one of the maintainers, just an avid follower :)