<?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>DEV Community: DewTM</title>
    <description>The latest articles on DEV Community by DewTM (@dewtm).</description>
    <link>https://dev.to/dewtm</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%2F3707420%2F9c6beaf5-89dc-4b35-a8a6-5e76b2b31aa1.PNG</url>
      <title>DEV Community: DewTM</title>
      <link>https://dev.to/dewtm</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/dewtm"/>
    <language>en</language>
    <item>
      <title>I built a spatial 3D knowledge graph with React Three Fiber &amp; Gemini AI 🌌</title>
      <dc:creator>DewTM</dc:creator>
      <pubDate>Mon, 12 Jan 2026 16:55:43 +0000</pubDate>
      <link>https://dev.to/dewtm/i-built-a-spatial-3d-knowledge-graph-with-react-three-fiber-gemini-ai-hpk</link>
      <guid>https://dev.to/dewtm/i-built-a-spatial-3d-knowledge-graph-with-react-three-fiber-gemini-ai-hpk</guid>
      <description>&lt;p&gt;Hey developers! 👋&lt;/p&gt;

&lt;p&gt;I've always felt that standard note-taking apps force our brains into unnatural, linear lists. Our thoughts are connected, chaotic, and spatial. So, I spent the last few weeks building an experiment to fix this.&lt;/p&gt;

&lt;p&gt;It's called &lt;strong&gt;Echo Archive&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;It's a spatial knowledge management system where your notes live in a 3D constellation.&lt;/p&gt;

&lt;h3&gt;
  
  
  🛠 The Tech Stack
&lt;/h3&gt;

&lt;p&gt;I wanted to keep it lightweight but visually immersive.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Frontend:&lt;/strong&gt; React 18 + Vite&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;3D Engine:&lt;/strong&gt; Three.js &amp;amp; React Three Fiber (drei)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;AI Core:&lt;/strong&gt; Google Gemini AI (via API)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Styling:&lt;/strong&gt; Tailwind CSS + Lucide Icons&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Storage:&lt;/strong&gt; LocalStorage (Privacy first, no cloud database)&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  🧠 The AI Integration
&lt;/h3&gt;

&lt;p&gt;The coolest part was integrating Gemini. Since I'm using the &lt;code&gt;gemini-2.5-flash&lt;/code&gt; model, I could implement:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt; &lt;strong&gt;Auto-Tagging:&lt;/strong&gt; The AI reads the note and suggests categories based on context.&lt;/li&gt;
&lt;li&gt; &lt;strong&gt;Smart Continue:&lt;/strong&gt; Writer's block? The AI finishes your sentence.&lt;/li&gt;
&lt;li&gt; &lt;strong&gt;Abstracts:&lt;/strong&gt; It automatically generates a TL;DR summary at the top of long notes.&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  🌌 The Challenge: Mobile 3D Navigation
&lt;/h3&gt;

&lt;p&gt;Making a 3D orbit control feel natural on a phone screen was tricky. I had to map touch events carefully to ensure you can pan the universe with one finger and zoom with two, without accidentally triggering the text editor.&lt;/p&gt;

&lt;h3&gt;
  
  
  🔗 Try it out
&lt;/h3&gt;

&lt;p&gt;I just released v1.0 and would love some feedback on the navigation feeling!&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Live Demo:&lt;/strong&gt; &lt;a href="https://echoarchive.tech" rel="noopener noreferrer"&gt;https://echoarchive.tech&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Let me know what you think about spatial interfaces. Are they the future or just a gimmick?&lt;/p&gt;

&lt;p&gt;Cheers!&lt;/p&gt;

</description>
      <category>react</category>
      <category>threejs</category>
      <category>javascript</category>
      <category>showdev</category>
    </item>
  </channel>
</rss>
