Zero-Shot Classification is the process of categorizing text or speech into labels that were not included during model training by using semantic reasoning rather than task-specific training data.
Voice AI platforms use Zero-Shot Classification to categorize customer conversations, detect emerging topics, identify intents, and automate routing for previously unseen requests.