“Jessie has observed that current voice-to-text models do not transcribe or understand children's voices as effectively as they do adult voices.”