video1.0<iframe src="https://www.loom.com/embed/c2cd9312c8574c2caf14840aff39a226" frameborder="0" width="1726" height="1294" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12941726Loomhttps://www.loom.com12941726https://cdn.loom.com/sessions/thumbnails/c2cd9312c8574c2caf14840aff39a226-8e929d131e0d44cf.gif483.662Hi, I am Adit, and this is my Applied AI Music Recommender final project for CodePath AI 1.1.0. I built an AI pipeline around my Module 3 rule based scoring engine with five steps: preference analysis, RAC knowledge base retrieval from custom genre and music theory files, unchanged numeric ranking of 28 songs, Claude rich explanations grounded in retrieved passages, and confidence scoring. My evaluation script ran 7 predefined profiles, all passed, with average confidence 0.54. I added guardrails for out of range energy like 1.5 and unknown genres. No action was requested from viewers.