<?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/23435db8309a450c91ef0c9eeca93246&quot; frameborder=&quot;0&quot; width=&quot;1986&quot; height=&quot;1489&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1489</height><width>1986</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1489</thumbnail_height><thumbnail_width>1986</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/23435db8309a450c91ef0c9eeca93246-c918e62a91dc05b0.gif</thumbnail_url><duration>507.719</duration><title>Vide Finder: AI Music Recommender Demo Pipeline 🎵</title><description>Hi, I am Youngmin Kim, and this Loom covers my project BDFinder, an AI Music Recommender built on the Cloud API. I explain the architecture for turning free text into game and mood analogy and a score, using my rule engine catalog and a nearly empty evaluator. I show live demos for happy pass and unhappy pass, with competence and energy results, and I demonstrate tuning with a few shot example. I also run pytest with 9 predefined and 5 normal profiles plus 2 categorical gap profiles, finishing in under 1 second. No action was requested from viewers.</description></oembed>