<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
     xmlns:atom="http://www.w3.org/2005/Atom"
     xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>GAIA Blog</title>
    <link>https://heygaia.io/blog</link>
    <description>Read the latest updates, insights, and stories from the GAIA team. Learn about AI, productivity, open-source development, and our journey building the future of personal AI assistants.</description>
    <language>en-us</language>
    <lastBuildDate>Tue, 03 Nov 2026 11:55:53 GMT</lastBuildDate>
    <atom:link href="https://heygaia.io/blog/rss.xml" rel="self" type="application/rss+xml"/>
    <image>
      <url>https://heygaia.io/images/logos/logo.webp</url>
      <title>GAIA Blog</title>
      <link>https://heygaia.io/blog</link>
    </image>
    
    <item>
      <title><![CDATA[Why Open Source?]]></title>
      <link>https://heygaia.io/blog/why-open-source</link>
      <guid isPermaLink="true">https://heygaia.io/blog/why-open-source</guid>
      <description><![CDATA[When we started building GAIA, one question kept coming up: why make it open source? The answer became clearer the more we thought about what kind of product and community we want to build.   Privacy and Trust  An AI assistant that integrates deeply with a user’s digital life has to take privacy seriously. People won’t trust something that has access to their emails, files, and schedules unless they can see how it actually works.  Open source creates transparency. Anyone can look at the code, v...]]></description>
      <pubDate>Sun, 02 Nov 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Aryan Randeriya]]></author>
      <category>Values</category>
      <enclosure url="https://heygaia.io/images/blog/open-source.webp" type="image/webp"/>
    </item>
    <item>
      <title><![CDATA[How we scaled GAIA to use thousands of tools]]></title>
      <link>https://heygaia.io/blog/how-tool-calling-works</link>
      <guid isPermaLink="true">https://heygaia.io/blog/how-tool-calling-works</guid>
      <description><![CDATA[We hit the wall at 200 tools.  That's when everything broke. The model would try to send a Slack message and call the Gmail API instead. Or it would freeze, paralyzed by choice, and refuse to call anything at all. Response times ballooned. Costs spiked. The context window was so bloated with tool schemas that the model barely had room to think.  OpenAI caps tools at around 128 per request. We weren't even close to that number technically, but functionally? The model was already drowning.  > How...]]></description>
      <pubDate>Sat, 15 Feb 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Dhruv Maradiya, Aryan Randeriya]]></author>
      <category>Tech</category>
      <enclosure url="https://heygaia.io/images/blog/how-tool-calling-works.webp" type="image/webp"/>
    </item>
    <item>
      <title><![CDATA[How We Migrated to a Monorepo using NX]]></title>
      <link>https://heygaia.io/blog/monorepo</link>
      <guid isPermaLink="true">https://heygaia.io/blog/monorepo</guid>
      <description><![CDATA[Changing Engines Mid-Flight  In the lifecycle of every high-growth startup, you hit a specific wall. It's the moment where your codebase transitions from "fast and loose" to "complex and fragile."  At GAIA, we hit that wall last week.  We are building a massive multi-surface platform: a Web app, a Desktop app, a Mobile app, a heavy-lifting Python backend, and a documentation site. Until recently, these lived in a flat, somewhat isolated structure. It worked for a prototype, but as we scaled, t...]]></description>
      <pubDate>Sun, 14 Dec 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Aryan Randeriya]]></author>
      <category>Engineering</category>
      <enclosure url="https://heygaia.io/images/blog/monorepo.webp" type="image/webp"/>
    </item>
    <item>
      <title><![CDATA[Introducing GAIA Desktop]]></title>
      <link>https://heygaia.io/blog/desktop</link>
      <guid isPermaLink="true">https://heygaia.io/blog/desktop</guid>
      <description><![CDATA[I love the web. It's where Gaia started, and it's where we iterate fastest. But there's a specific "stickiness" to a desktop app that a browser tab just can't match.  We built this desktop app for a very practical reason: we needed to use Gaia ourselves, every single day.  We wanted Gaia to be part of our daily workflows with zero friction. We needed to understand the product from our own experience as users—to feel the bugs, hit the limitations, and smooth out the rough edges. The only way to...]]></description>
      <pubDate>Thu, 04 Dec 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Aryan Randeriya]]></author>
      <category>Engineering</category>
      <enclosure url="https://heygaia.io/images/blog/desktop.webp" type="image/webp"/>
    </item>
    <item>
      <title><![CDATA[How does tool calling work under the hood?]]></title>
      <link>https://heygaia.io/blog/tool-calling-under-the-hood</link>
      <guid isPermaLink="true">https://heygaia.io/blog/tool-calling-under-the-hood</guid>
      <description><![CDATA[How Tool Call Works under the hood   What Is a Tool and Why Do We Need It?  A large language model LLM is just a text prediction engine. It can write code, summarize text, or plan actions — but it cannot _execute_ anything.  If you tell it to “send an email,” it can generate a perfect JSON body for that email, but it won’t actually send it.  That’s where tools come in.  A tool is a bridge between language and action — it’s a real function in your code that the model can ask you to execute.  Th...]]></description>
      <pubDate>Fri, 14 Nov 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Dhruv Maradiya]]></author>
      <category>Tech</category>
      <enclosure url="https://heygaia.io/images/blog/tool-calling-under-the-hood.webp" type="image/webp"/>
    </item>
    <item>
      <title><![CDATA[v0 - Public Beta]]></title>
      <link>https://heygaia.io/blog/public-beta</link>
      <guid isPermaLink="true">https://heygaia.io/blog/public-beta</guid>
      <description><![CDATA[We Finally Built the AI Assistant We Always Wanted  Most so-called "AI assistants" today don't feel like assistants at all. You talk to them, they answer something basic, and then leave the rest of the work to you. They don't remember anything. They don't understand context. They're not helpful in any serious way.  We wanted something different. Something personal. Something that doesn't wait around for instructions but actually takes initiative. An assistant that feels like it's working with y...]]></description>
      <pubDate>Sat, 02 Aug 2025 00:00:00 GMT</pubDate>
      <author><![CDATA[Aryan Randeriya, Dhruv Maradiya]]></author>
      <category>Product</category>
      <enclosure url="https://heygaia.io/images/blog/v0.webp" type="image/webp"/>
    </item>
  </channel>
</rss>