<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Nomic on Mathias WOLFF</title><link>https://www.blog-des-telecoms.com/en/tags/nomic/</link><description>Recent content in Nomic on Mathias WOLFF</description><generator>Hugo</generator><language>en-US</language><copyright>&lt;a href="https://www.blog-des-telecoms.com"&gt;Blog des télécoms&lt;/a&gt; © 2009 - 2026 by &lt;a href="https://www.linkedin.com/in/mathias-wolff-47a7941/"&gt;Mathias WOLFF&lt;/a&gt; is licensed under &lt;a href="https://creativecommons.org/licenses/by-nc-sa/4.0/"&gt;CC BY-NC-SA 4.0&lt;/a&gt;&lt;img src="https://mirrors.creativecommons.org/presskit/icons/cc.svg" style="max-width: 1em;max-height:1em;margin-left: .2em;"&gt;&lt;img src="https://mirrors.creativecommons.org/presskit/icons/by.svg" style="max-width: 1em;max-height:1em;margin-left: .2em;"&gt;&lt;img src="https://mirrors.creativecommons.org/presskit/icons/nc.svg" style="max-width: 1em;max-height:1em;margin-left: .2em;"&gt;&lt;img src="https://mirrors.creativecommons.org/presskit/icons/sa.svg" style="max-width: 1em;max-height:1em;margin-left: .2em;"&gt;</copyright><lastBuildDate>Mon, 06 Jul 2026 08:19:08 +0200</lastBuildDate><atom:link href="https://www.blog-des-telecoms.com/en/tags/nomic/index.xml" rel="self" type="application/rss+xml"/><item><title>When Switching Embeddings Doubles Recall: bge-m3 vs nomic on a Local RAG</title><link>https://www.blog-des-telecoms.com/en/blog/embedding-bge-m3-rag-local-wazo/</link><pubDate>Mon, 06 Jul 2026 09:00:00 +0200</pubDate><guid>https://www.blog-des-telecoms.com/en/blog/embedding-bge-m3-rag-local-wazo/</guid><description>&lt;p&gt;At Wazo, we run a fully local RAG. Data sovereignty is not a marketing argument: the knowledge base mixes product documentation, technical docs shared with our partners, open-source and proprietary code, and a large volume of private notes. All of it stays on our servers. Nothing goes to a cloud third party.&lt;/p&gt;
&lt;p&gt;That volume forces continuous improvement work. The RAG feeds both pre-sales and support — two use cases where a wrong or slow answer is costly. Recently, an embedding change took our search recall from &lt;strong&gt;0.45 to 0.95&lt;/strong&gt; and cut latency by &lt;strong&gt;twelve&lt;/strong&gt;. Measured, not guessed.&lt;/p&gt;</description></item></channel></rss>