Transcript of session - Custom media stream handling?

Does OpenVidu have a better way to implement transcription other than sending the datastreams from the client to two different endpoints?

Before implementing OpenVidu we used Kurento directly, to support transcription we sent audio streams to 2 endpoints, the Kurento server and our own transcription processing endpoint. With our cutover to OpenVidu Enterprise (with MediaSoup) we have replicated the same logic.
We don’t like the implementation, it means we are forcing the clients to send the data twice to the same datacenter.
What we would like to do is to have the client only send media streams to our OpenVidu server and have that either process transcription or forward the audio stream to our transcription endpoint.
Are there any hooks or interfaces we can implement to allow this to occur?
Is there a better way to handle this requirement under OpenVidu?

You can implement an special OpenVidu participant to work as a “transcriptor”. This participant will be subscribed to streams sent by real participants.

To do it, you can implement the OpenVidu protocol simultating a participant in a browser. The tricky part is that you will need a webrtc stack to receive the stream (but you already have one to receive the media from the browser). You can implement the transcriptor participant from scratch based on OpenVidu-protocol or you can reuse openvidu-browser implementation.

Best regards