A system for supervising an operator in a speech-driven environment includes an operator terminal configured for running a bi-directional speech task with an operator in a first spoken language wherein the speech task includes a plurality of successive nodes that are encountered by the operator, indicating progress through the task. A supervisor terminal is configured for running a bi-directional speech task with a supervisor in a second spoken language that is different from the first spoken language. A communication link exists between the operator terminal and supervisor terminal for communicating information therebetween. The operator terminal communicates, to the supervisor terminal, information regarding progress through the operator terminal speech task. The supervisor terminal progresses through the supervisor terminal speech task synchronized with the progress of the operator terminal speech task. A monitor is configured for displaying the operator's node-by-node progression upon receipt of speech data from said operator terminal.