<?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/caf1139a76194af8837b7b6f63946f65&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/caf1139a76194af8837b7b6f63946f65-3f74600c267754b7.gif</thumbnail_url><duration>211.776</duration><title>AI Triage for Bilingual Customer Emails 🚀</title><description>I shared my track case for the Mumswell AA Engineering intent world, focused on solving customer service email triage. I built an AI powered triage system that scores face confidence for structured outputs, drafts native language replacements in English and Arabic, and flags uncertain cases for human review. This was to reduce the operational cost of high CSC email volume and avoid quality loss and trust issues from simple translation. I used Python, Fast API, and PyTorch with open source models and posted the full code and trade offs on GitHub. No action was specifically requested from viewers.</description></oembed>