<?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: Jimmy Van Den Hoeck</title>
    <description>The latest articles on DEV Community by Jimmy Van Den Hoeck (@jimmy_vandenhoeck_5bcde).</description>
    <link>https://dev.to/jimmy_vandenhoeck_5bcde</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.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4226431%2F52a28b59-c017-415e-b162-dce29fe5f724.png</url>
      <title>DEV Community: Jimmy Van Den Hoeck</title>
      <link>https://dev.to/jimmy_vandenhoeck_5bcde</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/jimmy_vandenhoeck_5bcde"/>
    <language>en</language>
    <item>
      <title>I built a Windows trading app where you plug in any LLM (Grok, ChatGPT, Claude, Gemini, DeepSeek, Ollama)</title>
      <dc:creator>Jimmy Van Den Hoeck</dc:creator>
      <pubDate>Wed, 04 Nov 2026 14:28:30 +0000</pubDate>
      <link>https://dev.to/jimmy_vandenhoeck_5bcde/i-built-a-windows-trading-app-where-you-plug-in-any-llm-grok-chatgpt-claude-gemini-deepseek-13g</link>
      <guid>https://dev.to/jimmy_vandenhoeck_5bcde/i-built-a-windows-trading-app-where-you-plug-in-any-llm-grok-chatgpt-claude-gemini-deepseek-13g</guid>
      <description>&lt;p&gt;I'm the developer of Nova AI Trader, a Windows desktop app that trades on your own exchange account and lets you plug in your own AI model as extra advice per trade.&lt;/p&gt;

&lt;h2&gt;
  
  
  How it works
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;You connect your own exchange via API key. The app never holds your money.&lt;/li&gt;
&lt;li&gt;You pick a model: Grok, ChatGPT, Claude, Gemini, DeepSeek, or a local model through Ollama. API costs are on your own key.&lt;/li&gt;
&lt;li&gt;The model returns a strict JSON signal; position size, max loss and stop-loss are enforced in code, outside the model.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What was hard
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Rate limits&lt;/strong&gt; on both exchanges and LLM APIs, so we batch and cache market snapshots.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Prompt guardrails&lt;/strong&gt;: small models hallucinate tickers, so we use a fixed schema, an allowed-symbols list and temperature 0.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Exchange API quirks&lt;/strong&gt;: every exchange rounds lot sizes and reports errors differently.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;I'd love feedback, especially from people who run local models for agent-style decisions.&lt;/p&gt;

&lt;p&gt;Link: &lt;a href="https://www.novatradingapp.com/ai-trader?utm_source=devto" rel="noopener noreferrer"&gt;https://www.novatradingapp.com/ai-trader?utm_source=devto&lt;/a&gt; (paid app, €39/month or €390/year).&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Trading involves risk, you can lose your whole stake. Not financial advice.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>ai</category>
      <category>ollama</category>
      <category>crypto</category>
    </item>
  </channel>
</rss>
