Very interesting idea.

Would you be willing to share more info on the tech used in the process?

>I bought a couple of Chinese microphones

Which exact microphones? How long does their battery last?

As well as other parts of the process.

Not the OP but I've been tinkering with the same concept (24/7 processing).

'm using vosk browser: https://github.com/ccoreilly/vosk-browser

To do speech to text locally and it works very well for English.