<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Prompt-Engineering on jjshanks.net</title>
    <link>http://www.jjshanks.net/tags/prompt-engineering/</link>
    <description>Recent content in Prompt-Engineering on jjshanks.net</description>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Sun, 03 May 2026 09:00:00 -0800</lastBuildDate>
    <atom:link href="http://www.jjshanks.net/tags/prompt-engineering/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Building a Local OCR Pipeline for LLM Document Understanding</title>
      <link>http://www.jjshanks.net/posts/auto-ocr-pipeline/</link>
      <pubDate>Sun, 03 May 2026 09:00:00 -0800</pubDate>
      <guid>http://www.jjshanks.net/posts/auto-ocr-pipeline/</guid>
      <description>&lt;p&gt;&lt;strong&gt;Quick Take&lt;/strong&gt; I built a local OCR pipeline to convert technical documents into markdown that LLMs can actually reason from. The two biggest breakthroughs were cross-AI judging to break prompt-optimization deadlocks, and blank-region detection that increased figure recovery from 4 to 73 on a 72-page document.&lt;/p&gt;&#xA;&lt;p&gt;Recently, I started testing models and coding harnesses by giving them a technical write-up and asking them to build a learning site from it. The prompt I use:&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
