I thought this was "conversational" where the end user was speaking and the form was automatically getting filled up. This was something which we were actually looking to build as a use-case on top of our voice orchestration @ bolna.