Is there a way to build "tap to speak" inside of vapi? The problem we have is that we can't figure out how to get endpointing just the way we want it, and instead we'd rather just "Press to start recording" and then "Press to end recording" and ONLY have he bot respond once I click "Press to end recording". I'm a full stack web dev and I can implement things in the API but I can't think of a way to do this because it seems like by default, the bot HAS to response (the start speaking plan is a MAX of 5 seconds). Is there any way to do this?