Realtime computer vision prototyping and hosting framework for web browsers using WebRTC

Do you think it is useful if there is a framework to allow Machine Learning/Computer Vision developers to easily prototype and host their realtime video/audio models as WebApps,
where users can easily try the hosted models with realtime video/audio input and output on web browsers?
For example, with it, users can try a realtime face detection model with their WebCam inputs and see the result videos on web browsers in a real-time manner, without requirements for the MV/CV developers to know about WebApp and WebRTC development.

Though, there is already a great framework, Streamlit, for prototyping WebApps using ML/CV, it lacks realtime capability - users cannot send and receive realtime video/audio streams to/from the Streamlit app.
I think such a framework, something like Streamlit with realtime video/audio, is useful (at least for some ML/CV developers) and it's possible to develop it with WebRTC.

Do you think it's worth? If such a framework is available, will you use it?
Moreover, do you have any ideas to monetize it?