A voice server for hosting telephony services can include a stack interface configured to exchange telephony signaling information with a circuit-switched to packet-switched interface and a plurality of core clients, each having a call model stored therein. Each core client can be configured to select and execute at least one voice service associated with a call. The voice server further can include a speech processing system configured to convert speech to text and to convert text to speech for processing the call as well as a core processor. The core processor can be configured to distribute the telephony signaling information to the core clients and coordinate operation of the core clients and the speech processing system. The core processor also can exchange telephony signaling information with the circuit-switched to packet-switched interface via the stack interface.