<?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/9dc9826f731f416c90144d3ea776e7ee&quot; frameborder=&quot;0&quot; width=&quot;1662&quot; height=&quot;1246&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1246</height><width>1662</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1246</thumbnail_height><thumbnail_width>1662</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/9dc9826f731f416c90144d3ea776e7ee-38fd83adf1a05a1e.gif</thumbnail_url><duration>357.849</duration><title>Devon Orchestrator for GitHub Issue Planning</title><description>Hi, I am Daniel Rudd, and I am showing Devon as an issue orchestrator inside a GitHub issue. You can ask Devon to plan and iterate on a fix, then ask it to implement and raise a pull request, all with an audit trail and humans kept in the loop. I walk through an Apache Superset example where Devon plans a change after finding datetime.utc deprecation. Architecturally, GitHub webhooks hit a FastAPI service, identify intent, invoke Devon in CloudVM with durable sessions, and poll every 45 seconds. No viewer action was requested.</description></oembed>