<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Noel Moreno Lemus — Lab notes on applied AI</title><description>Notes on building agentic AI systems that hold up in production, by Noel Moreno Lemus.</description><link>https://nmlemus.github.io/</link><language>en</language><item><title>RAG End to End: A Practical Tutorial</title><link>https://nmlemus.github.io/blog/rag-end-to-end/</link><guid isPermaLink="true">https://nmlemus.github.io/blog/rag-end-to-end/</guid><description>What it is, how each piece works, how it is evaluated, a production example on Azure, and a comparison of Azure · Google Cloud · open source</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate><category>rag</category><category>retrieval</category><category>evaluation</category></item><item><title>The model decides what. Code decides how.</title><link>https://nmlemus.github.io/blog/the-model-decides-what/</link><guid isPermaLink="true">https://nmlemus.github.io/blog/the-model-decides-what/</guid><description>If AI writes excellent code, why turn every deterministic step of an agent into a tool? Because “sometimes” compounds.</description><pubDate>Wed, 09 Sep 2026 00:00:00 GMT</pubDate><category>agents</category><category>reliability</category></item><item><title>How Coding Harness Affects Token Consumption</title><link>https://nmlemus.github.io/blog/coding-harness-tokens/</link><guid isPermaLink="true">https://nmlemus.github.io/blog/coding-harness-tokens/</guid><description>Same model, same tasks, three harnesses. What Claude Code, OpenCode and Pi consumed, and how much my own configuration skewed the comparison.</description><pubDate>Mon, 07 Sep 2026 00:00:00 GMT</pubDate><category>coding-agents</category><category>cost</category></item><item><title>An Essential Guide to Safe Development Practices in the Era of Vibe Coding</title><link>https://nmlemus.github.io/blog/safe-vibe-coding/</link><guid isPermaLink="true">https://nmlemus.github.io/blog/safe-vibe-coding/</guid><description>Real failures from Replit and Gemini CLI, what research says about how autonomous agents fail, and the non-negotiable practices for AI-assisted development.</description><pubDate>Mon, 08 Dec 2025 00:00:00 GMT</pubDate><category>vibe-coding</category><category>safety</category></item></channel></rss>