<?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/de559bb0aef749559c79117b7f951250&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/de559bb0aef749559c79117b7f951250-1cf817707a1862dc.gif</thumbnail_url><duration>1361.773</duration><title>Unlocking the Power of TraceMind MCP Server for AI Optimization 🚀</title><description>Hello everyone, I&apos;m Kshitij from Mumbai India, and in this video, I introduced our TraceMind MCP server, designed to analyze real agent evaluation datasets for businesses looking to implement Gen AI. I discussed how our ecosystem addresses the need for deeper insights beyond standard leaderboards, allowing users to optimize models and improve performance. I showcased our four projects, including TraceVerde, a no-code instrumentation framework, and SmolTrace, an evaluation framework that generates custom datasets. I demonstrated the MCP server&apos;s capabilities, including analyzing leaderboards and estimating costs, and I encourage you to check out the TraceMind AI UI for more insights. If you find the MCP server useful, please give it a like!</description></oembed>