Website for audio classification

I want to create a website that would use my edge impulse audio classification model in real time (kind of like in this video. I also looked through this documentation but it seems like it only allows manually entering the features rather than automatically taking input. Is it possible to do this?

Here are some examples: