Integrate customer audio data for enhanced question answering

Expanding our question-answering capabilities to support audio-based data, not just text. The goal is to let users ask questions across all their content, including system audio, to extract insights no matter the format.

A recent user example showed the need to pull answers directly from recorded audio clips. Integrating that kind of data would unlock more complete analysis and remove the need to transcribe everything manually first.

Would love to see support for audio as a first-class data type in Q&A workflows.