An apparatus and method for monitoring an interaction between a caller and an automated voice response (AVR) system is provided. An audio communication from a caller is processed by executing an AVR script, which includes a plurality of instructions. A visual representation of the audio communication is presented substantially simultaneously with the audio communication to an agent based on the AVR script. The visual representation includes at least one field to be populated with information obtained from the caller and the information populated in the field can be updated by the agent.