Our predictive AI model is trained to recommend when a moderator might want to probe for more detail based on the syntax of a respondent’s answer.