Generally the present disclosure is directed to appliances that provide a user-specific response to a received voice command. In particular, the appliance can store a plurality of voice samples respectively associated with a plurality of users. The appliance can also store one or more preferences for each of the plurality of users. For example, the preferences can be input by the user and/or learned or inferred over time. When the appliance receives a human speech signal or voice command, it can match the received speech signal against one or more of the plurality of voice samples to identify the user. The preferences stored and associated with the identified user can then be obtained and the appliance can perform any requested operations in accordance with the obtained preferences. In such fashion, the appliance can provide a user-specific response to a received voice command.