<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0">
  <channel>
    <title>Jared Alonzo</title>
    <link>https://jaredalonzo.dev</link>
    <description>Engineering writing by Jared Alonzo.</description>
    <language>en</language>
    
    <item>
      <title><![CDATA[Four steps to cut LLM costs by 60–70%]]></title>
      <link>https://jaredalonzo.dev/posts/cost-optimization-four-moves</link>
      <guid>https://jaredalonzo.dev/posts/cost-optimization-four-moves</guid>
      <pubDate>Fri, 29 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[Measurement infrastructure, prompt caching, a deterministic prefilter, and a model swap. Each one validated against the previous baseline.]]></description>
    </item>
    <item>
      <title><![CDATA[How to Be the Person Everyone Trusts Across Teams]]></title>
      <link>https://jaredalonzo.dev/posts/cross-functional-trust</link>
      <guid>https://jaredalonzo.dev/posts/cross-functional-trust</guid>
      <pubDate>Wed, 27 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[Being trusted across functions isn't about being the smartest person in the room. It's about being the most legible one.]]></description>
    </item>
    <item>
      <title><![CDATA[What small wins actually do for a client relationship]]></title>
      <link>https://jaredalonzo.dev/posts/small-wins</link>
      <guid>https://jaredalonzo.dev/posts/small-wins</guid>
      <pubDate>Wed, 27 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[Client relationships aren't won in the big moments. They're sustained in the small ones — and most people don't figure that out until they need it.]]></description>
    </item>
    <item>
      <title><![CDATA[What I didn't know about MCP]]></title>
      <link>https://jaredalonzo.dev/posts/what-i-didnt-know-about-mcp</link>
      <guid>https://jaredalonzo.dev/posts/what-i-didnt-know-about-mcp</guid>
      <pubDate>Wed, 27 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[Adding career context to the agent looked like a few hours of work. It was, but not the hours I expected — MCP-bridged tools don't support sync invocation, and that one constraint touched every layer of the agent.]]></description>
    </item>
    <item>
      <title><![CDATA[Debugging a library you didn't write]]></title>
      <link>https://jaredalonzo.dev/posts/debugging-a-library-you-didnt-write</link>
      <guid>https://jaredalonzo.dev/posts/debugging-a-library-you-didnt-write</guid>
      <pubDate>Tue, 26 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[A diagnostic story: why a third-party library silently swallowed a 403, why verifying your assumptions saves a few headaches, and what it took to build a working live client from scratch.]]></description>
    </item>
    <item>
      <title><![CDATA[Two models, two jobs]]></title>
      <link>https://jaredalonzo.dev/posts/two-models-two-jobs</link>
      <guid>https://jaredalonzo.dev/posts/two-models-two-jobs</guid>
      <pubDate>Mon, 25 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[The classifier and the narrator have different jobs. They should have different prompts, different temperatures, and—it turned out—different models.]]></description>
    </item>
    <item>
      <title><![CDATA[Stateless graphs, stateful consumers]]></title>
      <link>https://jaredalonzo.dev/posts/game-context-tracker-langraph-state</link>
      <guid>https://jaredalonzo.dev/posts/game-context-tracker-langraph-state</guid>
      <pubDate>Sun, 24 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[Stateless graphs are easier to test. State has to live somewhere. The consumer is a natural home.]]></description>
    </item>
    <item>
      <title><![CDATA[Building a Real-Time Analyst Agent Using LangGraph]]></title>
      <link>https://jaredalonzo.dev/posts/building-a-real-time-analyst-agent-using-langchain</link>
      <guid>https://jaredalonzo.dev/posts/building-a-real-time-analyst-agent-using-langchain</guid>
      <pubDate>Thu, 21 May 2026 12:00:00 GMT</pubDate>
      <description><![CDATA[A walkthrough of building a queue-driven LangGraph agent that analyzes real-time events in sports.]]></description>
    </item>
  </channel>
</rss>