<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Mathieu Mafille</title><description>IT Specialist at Digilac writing about DevOps, AI tooling, and the Atlassian ecosystem. Open-source contributor and CS student based in Switzerland.</description><link>https://www.mafille.me/</link><item><title>Quoven — Building the Evidence Layer for Product Decisions</title><link>https://www.mafille.me/posts/2026/quoven-evidence-layer-for-product-decisions/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/quoven-evidence-layer-for-product-decisions/</guid><description>A build-in-public look at Quoven, my SaaS that turns a freeform brief into a sourced, structured dossier on a product decision — the multi-probe research pipeline, the stack, and the decisions I&apos;m proud of.</description><pubDate>Wed, 10 Jun 2026 07:00:00 GMT</pubDate><category>saas</category><category>ai</category><category>building-in-public</category><category>typescript</category><author>Mathieu Mafille</author></item><item><title>Claude Design — The First AI UI Tool That Doesn&apos;t Feel Like AI</title><link>https://www.mafille.me/posts/2026/claude-design-ai-native-ui-tool/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/claude-design-ai-native-ui-tool/</guid><description>A hands-on look at Claude Design: what it is, why Anthropic built it, the results that surprised me, the unmistakable feeling that a real designer was in the loop, and the two ways out — Figma to keep designing, or Claude Code to ship.</description><pubDate>Wed, 27 May 2026 07:00:00 GMT</pubDate><category>ai</category><category>design</category><category>anthropic</category><category>claude</category><category>ui</category><author>Mathieu Mafille</author></item><item><title>LLM Inference — Where Your VRAM Goes and How to Get It Back</title><link>https://www.mafille.me/posts/2026/llm-inference-optimization/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/llm-inference-optimization/</guid><description>A practical look at LLM inference: how context, KV cache, and model weights compete for VRAM, why Q8 on the KV cache is a free win, how quantization formats like GGUF and MLX compare, and why vLLM beats Ollama or LM Studio when you actually need throughput.</description><pubDate>Wed, 13 May 2026 07:00:00 GMT</pubDate><category>ai</category><category>llm</category><category>inference</category><category>performance</category><category>vllm</category><author>Mathieu Mafille</author></item><item><title>OpenClaw — An Open-Source Personal AI Agent You Actually Own</title><link>https://www.mafille.me/posts/2026/openclaw-personal-ai-agent/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/openclaw-personal-ai-agent/</guid><description>A hands-on look at OpenClaw: what it is, how MEMORY and SOUL give it a personality that sticks, why the brain is swappable but the architecture isn&apos;t, how to keep it secure, and why Telegram is the ideal gateway.</description><pubDate>Wed, 29 Apr 2026 07:00:00 GMT</pubDate><category>ai</category><category>open-source</category><category>productivity</category><category>telegram</category><author>Mathieu Mafille</author></item><item><title>BMAD Method — Structured AI Development From Idea to Code</title><link>https://www.mafille.me/posts/2026/bmad-method-structured-ai-development/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/bmad-method-structured-ai-development/</guid><description>A practical look at the BMAD method: what it is, how its specialized agents guide you through a real development workflow, why it pairs perfectly with Claude Code, and an honest comparison with vibe coding.</description><pubDate>Wed, 15 Apr 2026 07:00:00 GMT</pubDate><category>ai</category><category>productivity</category><category>open-source</category><author>Mathieu Mafille</author></item><item><title>Claude Code — The AI Coding Companion That Changed My Workflow</title><link>https://www.mafille.me/posts/2026/claude-code-ai-coding-companion/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/claude-code-ai-coding-companion/</guid><description>A deep dive into Claude Code: what it is, how it differs from Claude.ai, why the Max subscription is a bargain compared to API costs, and how MCP servers and built-in tools make it a genuine productivity multiplier.</description><pubDate>Wed, 01 Apr 2026 07:00:00 GMT</pubDate><category>ai</category><category>claude-code</category><category>productivity</category><category>mcp</category><author>Mathieu Mafille</author></item><item><title>Building an MCP Server for Switzerland&apos;s Public Procurement Platform</title><link>https://www.mafille.me/posts/2026/simap-mcp/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/simap-mcp/</guid><description>How I built simap-mcp, an open-source MCP server that lets AI assistants search and browse public tenders on SIMAP.ch.</description><pubDate>Wed, 18 Mar 2026 08:00:00 GMT</pubDate><category>mcp</category><category>ai</category><category>open-source</category><category>typescript</category><author>Mathieu Mafille</author></item><item><title>Hello World — Why I&apos;m Starting This Blog</title><link>https://www.mafille.me/posts/2026/hello-world/</link><guid isPermaLink="true">https://www.mafille.me/posts/2026/hello-world/</guid><description>A quick introduction to who I am, why I decided to start a blog, and what you can expect to find here.</description><pubDate>Mon, 16 Mar 2026 17:40:53 GMT</pubDate><category>personal</category><category>meta</category><author>Mathieu Mafille</author></item></channel></rss>