<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Predictiv Engineering perspectives</title><description>Field articles on applying AI to software delivery and support.</description><link>https://www.predictiv.dev/</link><language>en-gb</language><item><title>From instructions to enforced controls</title><link>https://www.predictiv.dev/perspectives/from-instructions-to-enforced-controls/</link><guid isPermaLink="true">https://www.predictiv.dev/perspectives/from-instructions-to-enforced-controls/</guid><description>A rule written in a prompt is a request. What changes when the rules an agent must follow are enforced where it acts, and which decisions stay human.</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate></item><item><title>Teaching agents an existing application</title><link>https://www.predictiv.dev/perspectives/teaching-agents-an-existing-application/</link><guid isPermaLink="true">https://www.predictiv.dev/perspectives/teaching-agents-an-existing-application/</guid><description>An agent new to a large, old codebase makes the mistakes a new hire makes, faster. What it needs to know before it changes anything.</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate></item><item><title>When coding is not the bottleneck</title><link>https://www.predictiv.dev/perspectives/when-coding-is-not-the-bottleneck/</link><guid isPermaLink="true">https://www.predictiv.dev/perspectives/when-coding-is-not-the-bottleneck/</guid><description>AI makes code cheap. What an organisation does next depends on whether it knows where its delivery really slows down.</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate></item></channel></rss>