video1.0<iframe src="https://www.loom.com/embed/2c63cc4fb32540149a814972895d4e12" frameborder="0" width="1166" height="874" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>8741166Loomhttps://www.loom.com8741166https://cdn.loom.com/sessions/thumbnails/2c63cc4fb32540149a814972895d4e12-50d90abc409b50fe.gif300.205Hi, in this Loom I show the project with the extended features. I start by installing requirements, then run the main code with default K=5, where the system ranks 5 songs with clear reasons like gender match, mood match, future closeness, and a confidence score where 1 is highest certainty. Next I run a custom profile, validate preferences within ranges, narrow to 9 candidates, then select the top 3 with a guardrail for diversity and ranking stability. I evaluate four Lofi and Tumor scenarios, get content scores from about 0.31 to 0.1 for certainty, and an average score around 0.78. I end by moving to an automated test pad, with no specific viewer action requested.