you would want to ask a question using any capture, and then add intent transitions. add words you think the user would put for each intent and then direct each intent to a different node