<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Ryan Mowlai — Blog</title><description>Technical deep-dives, paper summaries, and case studies.</description><link>https://rmowlai.com/</link><item><title>Offline Policy Evaluation Before You Touch the Plant</title><link>https://rmowlai.com/blog/offline-policy-evaluation-before-you-deploy/</link><guid isPermaLink="true">https://rmowlai.com/blog/offline-policy-evaluation-before-you-deploy/</guid><description>You cannot A/B test a water treatment plant. Here is how to decide whether an RL policy is worth deploying using only logged data from the incumbent controller.</description><pubDate>Mon, 15 Jun 2026 00:00:00 GMT</pubDate><category>Tutorial</category><category>Reinforcement Learning</category><category>Offline RL</category><category>Evaluation</category><category>Control</category></item></channel></rss>