<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Louis Planquart’s blog</title><link>https://louisplanquart.fr/en/blog/</link><description>Business applications, DevOps, systems integration, applied AI and SaaS products.</description><language>en-US</language><lastBuildDate>Wed, 29 Jul 2026 23:16:06 +0000</lastBuildDate><item><title>Patrick AI: the assistant that simplifies my team’s daily management</title><link>https://louisplanquart.fr/en/blog/patrick-ai-team-management-assistant/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/patrick-ai-team-management-assistant/</guid><pubDate>Wed, 29 Jul 2026 10:16:00 +0200</pubDate><description>How Patrick AI brings tasks, follow-ups, meeting notes and team monitoring together through a fully conversational interface.</description><category>Patrick IA</category><category>task management</category><category>team management</category><category>artificial intelligence</category><category>conversational assistant</category><category>Ollama</category><category>Local AI</category><category>automation</category><category>project monitoring</category><category>productivity</category><category>OpenProject</category><category>CRM</category></item><item><title>Ornith-1 stress test: is this 35B local model finally agentic-ready?</title><link>https://louisplanquart.fr/en/blog/ornith-1-35b-local-model-agentic-review/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/ornith-1-35b-local-model-agentic-review/</guid><pubDate>Mon, 27 Jul 2026 09:40:00 +0200</pubDate><description>I tested Ornith-1 35B locally on two RTX 3060 GPUs and a MacBook Pro M4: agentic coding, 64K context, VRAM limits and a comparison with GPT-5.6 SOL.</description><category>Ornith-1</category><category>Ollama</category><category>DeepReinforce</category><category>agentic coding</category><category>Local LLM</category><category>open source model</category><category>artificial intelligence</category><category>GGUF</category><category>llama.cpp</category><category>Codex</category><category>GPT-5.6 SOL</category><category>RTX 3060</category></item><item><title>PC SOFT crossed the line: charging developers’ end customers</title><link>https://louisplanquart.fr/en/blog/pc-soft-runtime-fee-end-customers/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/pc-soft-runtime-fee-end-customers/</guid><pubDate>Fri, 24 Jul 2026 14:00:00 +0200</pubDate><description>Why the announced fee on WINDEV applications deployed to end customers creates a major breach of trust with developers.</description><category>PC SOFT</category><category>WINDEV</category><category>WEBDEV</category><category>WINDEV Mobile</category><category>runtime royalty</category><category>software license</category><category>end users</category><category>Volaris Group</category><category>WLanguage</category><category>subscription</category><category>application migration</category><category>business software</category></item><item><title>IFS Cloud under the hood: exploring CRIM, APIs and the lifecycle</title><link>https://louisplanquart.fr/en/blog/ifs-cloud-crim-api-lifecycle/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/ifs-cloud-crim-api-lifecycle/</guid><pubDate>Thu, 23 Jul 2026 08:51:00 +0200</pubDate><description>Field notes on IFS Cloud 25R2: configurations, reports, interfaces, modifications, BPA, OData APIs, ACP, Build Place and Use Place.</description><category>IFS Cloud</category><category>CRIM</category><category>BPA</category><category>Workflow</category><category>API OData</category><category>API Explorer</category><category>IFS Connect</category><category>ACP</category><category>Build Place</category><category>Use Place</category><category>Developer Studio</category><category>ERP</category></item><item><title>AskLeopold: three months of vibe coding to frame projects better</title><link>https://louisplanquart.fr/en/blog/askleopold-ai-requirements-framing/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/askleopold-ai-requirements-framing/</guid><pubDate>Wed, 22 Jul 2026 08:46:00 +0200</pubDate><description>Jean-Francis Ochs and I built AskLeopold in three months to help business analysts frame requirements more effectively with AI.</description><category>AskLeopold</category><category>framing of the need</category><category>AMOA</category><category>artificial intelligence</category><category>vibe coding</category><category>expression of need</category><category>functional analysis</category><category>project management</category><category>software development</category><category>Jean-Francis Ochs</category></item><item><title>The power of vibe coding: six months that changed how I build software</title><link>https://louisplanquart.fr/en/blog/six-months-vibe-coding-software-development/</link><guid isPermaLink="true">https://louisplanquart.fr/en/blog/six-months-vibe-coding-software-development/</guid><pubDate>Tue, 21 Jul 2026 12:35:00 +0200</pubDate><description>After six months of vibe coding, I look back at my experiments, the limits of local AI and the first professional projects I migrated with Codex.</description><category>vibe coding</category><category>artificial intelligence</category><category>software development</category><category>Codex</category><category>Claude Code</category><category>GPT-5.6</category><category>Gitea</category><category>WinDev</category><category>application migration</category><category>Local AI</category></item></channel></rss>