<?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/453fdb90c5604a44bf7dbd36325bfb1b&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/453fdb90c5604a44bf7dbd36325bfb1b-15a82e437dbc0f54.gif</thumbnail_url><duration>132.945</duration><title>How Our Agentic Research System Works</title><description>This Loom explains the research agent and its supporting tools for live, web-based intelligent search. The speaker describes a web application where users can ask questions via a taskbar interface for real-time website and agentic search, with data transferred to a cloud AI by Anthropic for reasoning and source-based answers. They also mention ClickHouse as the database for real-time storage at production scale, including search history. Finally, they note additional capabilities such as directly using a tool called Twilio and using a Gleekhouse system for distributed compute and machine learning.</description></oembed>