video1.0<iframe src="https://www.loom.com/embed/5527f294cfe8456c88af9710d69766cc" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/5527f294cfe8456c88af9710d69766cc-5b0c9b77f3cd5904.gif183.019Hi everyone, I and my teammate Devanshu Murthy presented Shoescribe AI, a multi agent system that solves the challenge of writing high converting product descriptions for e commerce sellers. We bridge the research gap by taking a product name and category, then running a 4 agent pipeline that fetches real competitor data, extracts structured insights with category aware fallbacks and markdown stripping, writes short teaser and long form copy using the Groq powered Llama 3.18b, and finally scores it with a judge agent across five dimensions. We deployed it on Railway for fully automated public access. No action was requested from viewers.