Allow us to tune the denoising mode based on the node. In use cases where someone is talking to another person while on the call.. we want to have background noise removed.. but then when it comes at a data capture stage - it has bad accuracy due to denoising, it would be good to be able to tune the denoising based on the use case per node. Same thing with endpointing. When it comes to intent there is more context and even if there is a degree of inaccuracy it's not the end of the world. Wheras if you're doing data capture, you want it as accurate as possible even if that is trading a little bit of latecy. But right now i would have to make global changes.