<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<title>Pydantic on Pythonflow</title>
		<link>https://pythonflow.com/tags/pydantic/</link>
		<description>Recent content in Pydantic on Pythonflow</description>
		<generator>Hugo</generator>
		<language>en-us</language>
		
			<managingEditor>alicandonmez90@gmail.com (flowfelis)</managingEditor>
		
		
			<webMaster>alicandonmez90@gmail.com (flowfelis)</webMaster>
		
		
		
			<lastBuildDate>Sun, 21 Jun 2026 07:00:00 +0100</lastBuildDate>
		
			<atom:link href="https://pythonflow.com/tags/pydantic/index.xml" rel="self" type="application/rss+xml" />
			<item>
				<title>LangBot Day 5: Structured Output — Typed Responses with Pydantic</title>
				<link>https://pythonflow.com/posts/langbot-day-5-structured-output/</link>
				<pubDate>Sun, 21 Jun 2026 07:00:00 +0100</pubDate><author>alicandonmez90@gmail.com (flowfelis)</author>
				<guid>https://pythonflow.com/posts/langbot-day-5-structured-output/</guid>
				<description>&lt;p&gt;&lt;strong&gt;Recap:&lt;/strong&gt; Yesterday we wired LangBot&amp;rsquo;s prompt and model into one pipeline with LCEL&amp;rsquo;s pipe operator — &lt;code&gt;chain = prompt | llm&lt;/code&gt;. One &lt;code&gt;invoke()&lt;/code&gt; call now handles everything from formatting to generation. But we still get back a raw &lt;code&gt;AIMessage&lt;/code&gt; and have to reach into &lt;code&gt;.content&lt;/code&gt; for the text. Today LangBot returns real typed objects.&lt;/p&gt;&#xA;&lt;hr&gt;&#xA;&lt;h2 id=&#34;the-problem-unstructured-mush&#34;&gt;The problem: unstructured mush&lt;/h2&gt;&#xA;&lt;p&gt;Every LLM response so far has been an &lt;code&gt;AIMessage&lt;/code&gt; — a blob of text with no shape. To extract anything useful, you either parse it yourself or pray the model followed your formatting instructions. Both are fragile.&lt;/p&gt;</description>
			</item>
	</channel>
</rss>
