<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/"><channel><title>Dr. János Karancsi — Blog</title><description>Essays and posts on AI engineering, agentic systems, and data-driven analysis.</description><link>https://jkarancsi.com/</link><language>en-us</language><atom:link href="https://jkarancsi.com/rss.xml" rel="self" type="application/rss+xml"/><item><title>What Legacy Code Taught Me About Learning AI From Scratch</title><link>https://jkarancsi.com/blog/what-legacy-code-taught-me-about-learning-ai-from-scratch/</link><guid isPermaLink="true">https://jkarancsi.com/blog/what-legacy-code-taught-me-about-learning-ai-from-scratch/</guid><description>LinkedIn post on what CERN legacy code taught me about durability, and building a transformer from scratch to learn AI fundamentals.</description><pubDate>Sun, 05 Jul 2026 00:00:00 GMT</pubDate></item><item><title>LinkedIn post - AI should navigate knowledge, not rediscover it</title><link>https://jkarancsi.com/blog/post-ai-knowledge-base/</link><guid isPermaLink="true">https://jkarancsi.com/blog/post-ai-knowledge-base/</guid><description>Published post on the agentic knowledge-base approach and context discipline.</description><pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate></item><item><title>COVID-19 super-spreaders - a data-driven analysis (2020)</title><link>https://jkarancsi.com/blog/covid-superspreader-analysis/</link><guid isPermaLink="true">https://jkarancsi.com/blog/covid-superspreader-analysis/</guid><description>Public analysis arguing super-spreader events break SEIR models; a physicist&apos;s data-driven take.</description><dc:date>2020</dc:date></item></channel></rss>