<?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/0f9b0e8517bd45f0b4165c04e4329c2b&quot; frameborder=&quot;0&quot; width=&quot;1916&quot; height=&quot;1437&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1437</height><width>1916</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1437</thumbnail_height><thumbnail_width>1916</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/0f9b0e8517bd45f0b4165c04e4329c2b-6cc7f32a1ee206ed.gif</thumbnail_url><duration>524.026</duration><title>How a Meeting AI Creates Accountable Records</title><description>This Loom explains a meeting intelligence setup to fix Northstar’s unaccountable meeting actions and missing memory across 120 weekly meetings. The system uses a consistent taxonomy with five meeting types and a minimum record standard, an assistant that drafts records from messy transcripts while refusing to guess missing owners or dates, and a register plus dashboard to surface overdue and stuck items. Managers can view action status in one screen, including counts such as six overdue actions and four stuck actions. Results from the pilot show minutes drafted under five minutes, over 90 percent of actions with an owner and date, about 80 percent less post meeting admin, and an estimated savings of over 2,500 hours per year.</description></oembed>