An application may be hosted for utilization by a remote computing platform. User interface (UI) elements of a UI generated by the hosted application may be identified. Proxy UI elements may be generated. Each of the proxy UI elements may correspond to one or more of the identified UI elements. A transcript of an audio sample may be processed. The audio sample may comprise an utterance of a user of the remote computing platform. The transcript of the audio sample may comprise at least one word corresponding to one or more of the proxy UI elements. A functionality of the hosted application may be invoked. The invoked functionality may correspond to one or more of the UI elements corresponding to the one or more of the proxy UI elements.