#MASSIVE

'A practical framework for evaluating modern voice agents that extends beyond ASR and WER to include task success, barge-in handling, hallucination-under-noise, safety, and perceptual quality.'