<?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/bf19084106064101a984bf4645830a2e&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/bf19084106064101a984bf4645830a2e-fff2f1bb7eddd4bb.gif</thumbnail_url><duration>219.427</duration><title>Music Recommender User Profiles and Guardrails</title><description>I added a user profile generator to my music recommender. Using a Streamlit front end, users click through a list of songs and mark which they like or dislike, and the system generates a profile using heuristics like pop, high energy, and happy, then produces top recommendations based on that. I also demonstrated terminal guardrails to sanity check the backend outputs, like making sure recommendations are in descending order and genre constraints are respected. Next, I want to build a time gated profile using your last seven weeks of listens, and weight recent activity at about 60 percent. No action is requested from viewers.</description></oembed>