<?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: Yuvilen Naidoo</title>
    <description>The latest articles on Forem by Yuvilen Naidoo (@yuvilen_naidoo_c9420c472b).</description>
    <link>https://forem.com/yuvilen_naidoo_c9420c472b</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%2F3702235%2Fd393a2b0-8215-4912-936e-5500610d0740.png</url>
      <title>Forem: Yuvilen Naidoo</title>
      <link>https://forem.com/yuvilen_naidoo_c9420c472b</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://forem.com/feed/yuvilen_naidoo_c9420c472b"/>
    <language>en</language>
    <item>
      <title>Why I built a WhatsApp-style AI companion because existing apps felt too fake</title>
      <dc:creator>Yuvilen Naidoo</dc:creator>
      <pubDate>Fri, 09 Jan 2026 11:00:06 +0000</pubDate>
      <link>https://forem.com/yuvilen_naidoo_c9420c472b/why-i-built-a-whatsapp-style-ai-companion-because-existing-apps-felt-too-fake-40ea</link>
      <guid>https://forem.com/yuvilen_naidoo_c9420c472b/why-i-built-a-whatsapp-style-ai-companion-because-existing-apps-felt-too-fake-40ea</guid>
      <description>&lt;h2&gt;
  
  
  The "Uncanny Valley" Problem
&lt;/h2&gt;

&lt;p&gt;I've been testing a lot of AI companion apps recently (for research, obviously 👀), and I noticed a recurring problem: &lt;strong&gt;Gamification kills immersion.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Most apps like Candy.ai or Character.ai force you to interact with a 3D avatar, navigate "gacha" style menus, or collect coins. Every time I saw a coin pop up, my brain was reminded: &lt;em&gt;"This is a video game. This isn't real."&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;I wanted something that felt &lt;strong&gt;grounded&lt;/strong&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  The Solution: UI Psychology
&lt;/h2&gt;

&lt;p&gt;I realized that for most of us, our "real" relationships happen in one specific place: &lt;strong&gt;The DMs.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;We don't talk to our partners via 3D floating heads; we talk to them via WhatsApp, iMessage, or Telegram. So, I decided to build &lt;strong&gt;Amio&lt;/strong&gt; (&lt;a href="https://meetamio.ai" rel="noopener noreferrer"&gt;https://meetamio.ai&lt;/a&gt;) with a strict design philosophy: &lt;strong&gt;Psychological Realism.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F764exvirbesg7geb4shx.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F764exvirbesg7geb4shx.png" alt="1-1 chats feel natural and personas carry into group chats" width="800" height="421"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  The Tech Stack 🛠️
&lt;/h2&gt;

&lt;p&gt;I built this as a PWA (Progressive Web App) to mimic the mobile experience perfectly.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Frontend:&lt;/strong&gt; React (for that snappy, app-like feel).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The Brains:&lt;/strong&gt; I'm routing requests to &lt;strong&gt;Sao10k: Llama 3 Euryale 70B v2.1&lt;/strong&gt; via OpenRouter.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Why Euryale?&lt;/strong&gt; In my benchmarking, it beats GPT-4o for creative, open-ended storytelling and "EQ" (Emotional Intelligence). It doesn't have the "Corporate HR" filter that Claude or Gemini have.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Key Features for Realism
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt; &lt;strong&gt;No Avatars:&lt;/strong&gt; Just a profile pic and text. Your imagination does the rest.&lt;/li&gt;
&lt;li&gt; &lt;strong&gt;Latency Injection:&lt;/strong&gt; I actually added artificial "typing" delays. Real humans don't reply in 0.2ms.&lt;/li&gt;
&lt;li&gt; &lt;strong&gt;Intensity Slider:&lt;/strong&gt; You can set the AI to be "Reserved" (shy) or "Bold" (proactive).&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Looking for Feedback
&lt;/h2&gt;

&lt;p&gt;I'm currently running a &lt;strong&gt;Beta&lt;/strong&gt; to stress-test the context window and the Llama 3 inference speeds.&lt;/p&gt;

&lt;p&gt;If you want to try "breaking" it or just see if the WhatsApp UI tricks your brain better than the Avatar apps, give it a shot.&lt;/p&gt;

&lt;p&gt;👉 &lt;strong&gt;Live Demo:&lt;/strong&gt; &lt;a href="https://meetamio.ai" rel="noopener noreferrer"&gt;https://meetamio.ai&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;I'd love to hear your thoughts on the UI choice—does the "text-only" approach feel more real to you, or do you prefer the 3D avatars? Let me know in the comments!&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>ai</category>
      <category>react</category>
      <category>webdev</category>
    </item>
  </channel>
</rss>
