<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Marius Argatu · Blog</title><description>Long-form technical writing on software testing, with a focus on AI and LLM evaluation.</description><link>https://www.mariusargatu.com/</link><item><title>Your Evals Are Checks, Not Tests</title><link>https://www.mariusargatu.com/blog/your-evals-are-checks-not-tests/</link><guid isPermaLink="true">https://www.mariusargatu.com/blog/your-evals-are-checks-not-tests/</guid><description>Air Canada&apos;s chatbot cost CAD $812 for an answer evals scored as faithful. Five classical testing patterns catch what your eval dashboard cannot.</description><pubDate>Thu, 11 Jun 2026 00:00:00 GMT</pubDate><category>llm</category><category>evals</category><category>rag</category><category>testing</category><category>agentic</category></item></channel></rss>