As provided herein, an audio signal (e.g., a user's spoken statement “check email”) may be received from the user of a client device accessing a webpage (e.g., a webmail webpage) utilizing a browser. The webpage may be parsed, analyzed, and/or mapped to identify one or more elements on the webpage (e.g., a send/receive email user interface element). The audio signal may be evaluated relative to the one or more elements on the webpage (e.g., such as by a dictionary comprising one or more audio signals paired to one or more audio commands) to identify an audio command. The audio command may be transformed into a text command (e.g., a script corresponding to a script action that invokes the send/receive email user interface element). The text command may be utilized to invoke an event (e.g., send/receive email) corresponding to an element on the webpage.