Possibilities of a public, DeepSpeech-powered web app transcription tool?

Stopping by from the world of audio production and as a fan of open source tech (though a complete amateur as far as programming goes)… I was just curious: Has anyone attempted to create a public, open source web transcriber using this technology? Or, what would such a project require?

I see a few GitHub projects that look somewhat similar, and of course articles about building apps locally on one’s computer, but nothing quite what I imagine, which would be a simple web app anyone could use: A site where one could upload a sound file, toggle a few options (whether to filter out filler words like “um” and “uh,” whether to have timecodes, etc), and it would spit out the transcribed text.

There are so many companies making money off this sort of tech (e.g. Rev, Trint, Temi) but nothing free and open source. Pardon my programming illiteracy, but I am curious what this community thinks. You don’t know how many would be thankful for such a tool - not least radio producers!

It is expensive to run cloud transcription service. Cloud CPU costs a lot. Less than charges from speech companies like Rev but still a lot. You’d better run transcription locally on your computer.

Ah I see - I admit I wasn’t thinking about that, but can see how that would be a big hurdle. I guess I’ll try and figure out how to run transcription locally!