Question

A field service team records short voice memos describing equipment issues. Your team wants to build a lightweight app in Foundry that accepts each audio recording, understands the spoken content, and returns a written troubleshooting suggestion in one workflow. Which approach best meets this requirement with the least added complexity?

Accepted Answer

Deploy a deployed multimodal model that accepts audio input and returns a text response

Answer

Deploy a text-only chat model and require staff to type transcripts of each memo manually

Answer

Use an image-generation model to visualize the equipment described in the audio

Answer

Use Azure Content Understanding to extract predefined fields from the audio into a structured schema