video1.0<iframe src="https://www.loom.com/embed/9de1e0c8eece4d65a46874497defe244" frameborder="0" width="1730" height="1297" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12971730Loomhttps://www.loom.com12971730https://cdn.loom.com/sessions/thumbnails/9de1e0c8eece4d65a46874497defe244-941024f99c519622.gif406.461Hey everyone, I'm excited to introduce our new AI agent designed to help analyze large trace files, which can often be overwhelming. This agent can run queries on trace files like J, Profero Chrome, or Ancestrap, allowing us to identify patterns and details about kernel types and groups. In this video, I demonstrate how to use the agent with a 200-300 MB trace file, breaking down its contents and mapping kernels to LLM inference operations. I encourage you to follow along as I show how to distill complex data into actionable insights.