<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Posts on Letters to My Future Self</title>
    <link>https://vladimir.shulyak.net/posts/</link>
    <description>Recent content in Posts on Letters to My Future Self</description>
    <image>
      <title>Letters to My Future Self</title>
      <url>https://vladimir.shulyak.net/papermod-cover.png</url>
      <link>https://vladimir.shulyak.net/papermod-cover.png</link>
    </image>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Fri, 14 Aug 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://vladimir.shulyak.net/posts/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Agentic Data Science: Gates, Interfaces, and Discipline</title>
      <link>https://vladimir.shulyak.net/posts/agentic-data-science-gates/</link>
      <pubDate>Fri, 14 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://vladimir.shulyak.net/posts/agentic-data-science-gates/</guid>
      <description>&lt;p&gt;The more work I delegated to agents, the more attention I had to spend protecting my own agency.&lt;/p&gt;
&lt;p&gt;I have been testing a more automated way of doing data science. In the morning, I define the larger question, hypotheses, and possible paths. Then the agents run experiments, sometimes on a cluster or GPU, while I go to meetings. I check in, answer questions, and redirect when needed. By evening, I have results. The next morning, I review them and repeat.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
