<?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/af64b030839044aa8c536b91de93d634&quot; frameborder=&quot;0&quot; width=&quot;1920&quot; height=&quot;1440&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1440</height><width>1920</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1440</thumbnail_height><thumbnail_width>1920</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/af64b030839044aa8c536b91de93d634-988646639d873872.gif</thumbnail_url><duration>95.5</duration><title>Build and Judge AI Agents With Profiles</title><description>This Loom demonstrates creating and running a turn-based game using “Astra” agents powered by real linking profiles and deep-researched persona data. The creator adds a person by providing a real linking profile, then Astra generates a map and builds a persona including taste, state, and background based on the provided information. After setup, the user can create multiple builds and judges and start the turn-based game, taking actions each round such as “research the voice base.” When finished, the judges provide a transcript of different people’s artifacts as a demo of the results.</description></oembed>