<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Forem: olivier delfosse</title>
    <description>The latest articles on Forem by olivier delfosse (@olivier_delfosse_6df9f498).</description>
    <link>https://forem.com/olivier_delfosse_6df9f498</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3572776%2F975e7281-23b7-4317-83e9-16c17e6eabd2.jpg</url>
      <title>Forem: olivier delfosse</title>
      <link>https://forem.com/olivier_delfosse_6df9f498</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://forem.com/feed/olivier_delfosse_6df9f498"/>
    <language>en</language>
    <item>
      <title>I built a Next.js + AI app that turns any book into a character map</title>
      <dc:creator>olivier delfosse</dc:creator>
      <pubDate>Sat, 18 Oct 2025 16:34:26 +0000</pubDate>
      <link>https://forem.com/olivier_delfosse_6df9f498/i-built-a-nextjs-ai-app-that-turns-any-book-into-a-character-map-49h0</link>
      <guid>https://forem.com/olivier_delfosse_6df9f498/i-built-a-nextjs-ai-app-that-turns-any-book-into-a-character-map-49h0</guid>
      <description>&lt;p&gt;&lt;strong&gt;Why I built it:&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I love complex stories, but I always forget who’s who when I come back to a book.&lt;br&gt;
So I built Booklaxy, a web app that generates interactive character maps and spoiler-free summaries using AI.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;What it does&lt;/strong&gt;    &lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;🧩 AI-generated summaries per chapter&lt;/li&gt;
&lt;li&gt;🌌 Visual character relationships (interactive graph)&lt;/li&gt;
&lt;li&gt;📝 Personal notes and book wiki&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;👉 Try it here → booklaxy.com&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;How I built it&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Tech stack:&lt;br&gt;
Next.js 14, MongoDB Atlas, Gemini AI, Tailwind CSS, GCP cloud run&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;What I learned&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Gemini handles multilingual books well, but prompt updates can break old results =&amp;gt; always test to avoid regressions! (more complex that the basic tests)&lt;/li&gt;
&lt;li&gt;AI integration with the App is always heavier than it looks: design, async logic and retries add unexpected complexity.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Looking for feedback&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I’d love to get feedback from fellow builders:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;How would you improve the onboarding UX?&lt;/li&gt;
&lt;li&gt;For those who’ve built prompt-based AI APIs, any tricks to keep things stable and efficient? :) &lt;/li&gt;
&lt;/ul&gt;

</description>
      <category>nextjs</category>
      <category>ai</category>
      <category>sideprojects</category>
      <category>d3js</category>
    </item>
  </channel>
</rss>
