video1.0<iframe src="https://www.loom.com/embed/e04e18e0b9ca49229e3ed84ed7821335" frameborder="0" width="1728" height="1296" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12961728Loomhttps://www.loom.com12961728https://cdn.loom.com/sessions/thumbnails/e04e18e0b9ca49229e3ed84ed7821335-a152fbbd614934ee.gif137.49This Loom shows how to improve an AI agent using real call transcripts. The presenter reviews a specific call in Copilot, then reduces repetitive explanations by changing the agent to mention an option only once and to ask a short confirmation or combination question when the customer response is ambiguous. They also demonstrate a broader approach by having Copilot review the last five or last 10 transcripts from the last month to surface patterns and key findings. Finally, they show how to review and test proposed edits in the agent builder using the Canvas UI.