We are creating a voice assistant application. We want the application to be always listening, but only record when the user talks to it, stops recording when the user finish talking, and save that audio (to send it via an API later to an AI platform, but that's not part of this).
To be clear, we don't want to press a button to record, or esle there will be no need for this project. What we want is the recording triggered to start once the web application is addressed, stop recording by itself when thr user stops talking and save the wave file. We've seen possible solutions to make that happen, but we will want to hear from you before sharing suggestions. Possibly, key there is for the system to start recording by comparing the background standard voice level to the newly higher wave that occurs when a speaker starts talking to the mic.
Your solution could be a script that can run within a web application, an open source platform you know and we can integrate to our system (of course you get paid for introducing us to that), or even a Unity web blank project that does just that.
I think I can do what you tell me, I have audio skills, I can also clear sound, I have experience in making audio for children's films, games, And others