<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/f8ccb63298174e329680dcadf8459623&quot; frameborder=&quot;0&quot; width=&quot;1460&quot; height=&quot;1095&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1095</height><width>1460</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1095</thumbnail_height><thumbnail_width>1460</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/f8ccb63298174e329680dcadf8459623-375d6d2b7f669a0a.gif</thumbnail_url><duration>206.806</duration><title>Designing a One-Tap Agent Feedback UX</title><description>This Loom walks through a one tap UX feedback feature for World AI that lets users tell agents what and why something went wrong. It shows an expense system issue that silently drops records, provides positive and negative buttons, and then offers four pre written responses including access to the receipt and escalation context (the seventh time it has happened this week, to escalate next week). The creator explains they chose closed ended buttons over open text because open ended surveys have a 18% skip rate versus 1 to 2% for closed ended prompts, and emphasizes that inline feedback increases effectiveness. They also include a skip option and discuss storing both correct and incorrect feedback.</description></oembed>