<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Opportune Blog</title><description>Opinion, deep dives, and learning notes from the Opportune engineering team.</description><link>https://51474130.landing-9h2.pages.dev/</link><item><title>Context engineering takes 6 days, not 6 months</title><link>https://51474130.landing-9h2.pages.dev/blog/context-engineering-6-days/</link><guid isPermaLink="true">https://51474130.landing-9h2.pages.dev/blog/context-engineering-6-days/</guid><description>[Deep dive] Most teams treat production LLM failures as an iteration problem. They&apos;re diagnosis problems. Here&apos;s what changes when you stop iterating and start diagnosing.</description><pubDate>Tue, 12 May 2026 00:00:00 GMT</pubDate><category>deep-dive</category><category>context-engineering</category><category>llm-reliability</category><category>production-ai</category></item><item><title>Why we&apos;re writing</title><link>https://51474130.landing-9h2.pages.dev/blog/welcome/</link><guid isPermaLink="true">https://51474130.landing-9h2.pages.dev/blog/welcome/</guid><description>[Opinion] An opening note on why Opportune is publishing — and what to expect from the three streams: opinion, deep dives, and learning.</description><pubDate>Tue, 05 May 2026 00:00:00 GMT</pubDate><category>opinion</category><category>intro</category></item></channel></rss>