Abstract: A system and method are described for processing a spoken request from a user. In one embodiment, a method is disclosed for attempting to recognize a spoken request from a user with a speech recognition engine above a predetermined level of accuracy. If the spoken request is not recognized above the predetermined level of accuracy, the spoken request is provided to a level one agent. If the level one agent does not recognize the request, a voice connection is established between the user and a level two agent. In another embodiment, a method is disclosed for determining whether a silent response system recognizes a spoken request from a user above a predetermined level of accuracy. A response is provided to the user if the silent response system recognizes the spoken request. Otherwise, a voice connection is established between the user and a call center.