<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Seyn — Writing</title>
    <link>https://seynlabs.com/blog</link>
    <atom:link href="https://seynlabs.com/blog/rss.xml" rel="self" type="application/rss+xml" />
    <description>Essays and field notes from the Seyn team on reading how businesses actually run, extracting ground truth, and proving where AI pays.</description>
    <language>en-us</language>
    <lastBuildDate>Mon, 17 Aug 2026 09:00:00 GMT</lastBuildDate>
    <item>
      <title>Don&apos;t automate a broken process</title>
      <link>https://seynlabs.com/blog/dont-automate-a-broken-process</link>
      <guid isPermaLink="true">https://seynlabs.com/blog/dont-automate-a-broken-process</guid>
      <pubDate>Mon, 17 Aug 2026 09:00:00 GMT</pubDate>
      <category>Method</category>
      <description>AI on a broken process buys you faster mistakes. One walk, three questions, three piles: a field method for finding the work worth automating.</description>
    </item>
    <item>
      <title>The $200K project failed. The $20K project worked. Same company.</title>
      <link>https://seynlabs.com/blog/the-200k-project-failed-the-20k-project-worked</link>
      <guid isPermaLink="true">https://seynlabs.com/blog/the-200k-project-failed-the-20k-project-worked</guid>
      <pubDate>Mon, 20 Jul 2026 09:00:00 GMT</pubDate>
      <category>Field notes</category>
      <description>We went back through a year of conversations and coded them. 86 mid-market operators, counted. Why the cheap, small, slightly embarrassing thing keeps beating the expensive, ambitious one.</description>
    </item>
    <item>
      <title>Why we read before we build</title>
      <link>https://seynlabs.com/blog/why-we-read-before-we-build</link>
      <guid isPermaLink="true">https://seynlabs.com/blog/why-we-read-before-we-build</guid>
      <pubDate>Tue, 12 May 2026 09:00:00 GMT</pubDate>
      <category>Method</category>
      <description>Most AI projects start from a workshop and a wish list. We start from the evidence already sitting in your systems — and that changes what gets built.</description>
    </item>
  </channel>
</rss>
