<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>Agentic PM</title>
<link>https://agentic-pm.ai/blog</link>
<description>Writing on commitment drift, agentic product management, and local-first AI tooling.</description>
<language>en</language>
<atom:link href="https://agentic-pm.ai/blog/rss.xml" rel="self" type="application/rss+xml"/>
<item>
<title>We ran a commitment ledger over 15 real meetings. It found four failures the note taker missed.</title>
<link>https://agentic-pm.ai/blog/commitment-drift-15-real-meetings</link>
<guid isPermaLink="true">https://agentic-pm.ai/blog/commitment-drift-15-real-meetings</guid>
<pubDate>Wed, 09 Sep 2026 09:00:00 GMT</pubDate>
<description>A commercial note taker summarised every meeting and marked every action item Pending. Comparing meetings to each other found four commitments that had quietly died.</description>
</item>
<item>
<title>Why your meeting notes structurally cannot catch a dropped commitment</title>
<link>https://agentic-pm.ai/blog/why-your-meeting-notes-cannot-catch-drift</link>
<guid isPermaLink="true">https://agentic-pm.ai/blog/why-your-meeting-notes-cannot-catch-drift</guid>
<pubDate>Wed, 09 Sep 2026 09:00:00 GMT</pubDate>
<description>This is not a quality problem with note takers. It is a shape problem. A per-meeting artefact cannot represent something that stopped happening.</description>
</item>
<item>
<title>Local-first AI for product teams, and why your security review keeps saying no</title>
<link>https://agentic-pm.ai/blog/local-first-ai-tools-for-product-teams</link>
<guid isPermaLink="true">https://agentic-pm.ai/blog/local-first-ai-tools-for-product-teams</guid>
<pubDate>Wed, 09 Sep 2026 09:00:00 GMT</pubDate>
<description>Most AI meeting tools upload your calls to a vendor's cloud. For regulated teams and most of Europe, that is where the conversation ends. Here is the alternative shape.</description>
</item>
</channel>
</rss>