Âm thanh chia sẻ

Next is Structured Output. This helps format the agent's response in a defined structure. Once enabled, you can select the taxonomy, choose the required labels using Select All Labels, and map it to the right process unit. If needed, you can also enable reasoning to capture the agent's thought process, include a confidence score, and add references to show the source of the response. This ensures that even multi-agent setup execution remains the predictable and aligned business goals.