<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Function Calling on Gabriel Mongeon</title><link>https://gabrielmongeon.ca/en/tags/function-calling/</link><description>Recent content in Function Calling on Gabriel Mongeon</description><generator>Hugo</generator><language>en</language><lastBuildDate>Fri, 15 May 2026 00:06:00 -0400</lastBuildDate><atom:link href="https://gabrielmongeon.ca/en/tags/function-calling/index.xml" rel="self" type="application/rss+xml"/><item><title>Function Calling: Teaching Tools to the Assistant</title><link>https://gabrielmongeon.ca/en/2026/05/raspberry-pi-voice-assistant-function-calling/</link><pubDate>Fri, 15 May 2026 00:06:00 -0400</pubDate><guid>https://gabrielmongeon.ca/en/2026/05/raspberry-pi-voice-assistant-function-calling/</guid><description>&lt;p>&lt;em>This post is part of the &lt;a href="https://gabrielmongeon.ca/en/series/voice-assistant-on-raspberry-pi/">Voice Assistant on Raspberry Pi&lt;/a> series.&lt;/em>&lt;/p>
&lt;p>In article #5, we injected weather data into every conversation, even for questions like &lt;em>&amp;ldquo;what&amp;rsquo;s your name?&amp;rdquo;&lt;/em>. That wastes tokens. Function calling fixes this: the LLM decides when it needs a tool and only calls it when the question actually warrants it.&lt;/p>
&lt;p>The complete code for this article is available on &lt;a href="https://github.com/mongeon/code-examples/tree/main/dotnet/ai/audio-assistant/06-function-calling">GitHub&lt;/a>.&lt;/p></description></item></channel></rss>