<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>Fathy Gad — Insights</title>
  <link>https://fathygad.me/insights/</link>
  <atom:link href="https://fathygad.me/insights/feed.xml" rel="self" type="application/rss+xml"/>
  <description>Design automation, computational engineering and governed AI in AEC.</description>
  <language>en</language>
  <item><title>Automation is a bidding advantage, not just a time saving</title><link>https://fathygad.me/insights/automation-is-a-bidding-advantage/</link><guid>https://fathygad.me/insights/automation-is-a-bidding-advantage/</guid><pubDate>Sun, 25 Oct 2026 08:00:00 GMT</pubDate><description>Hours returned by automation only create value when they are redeployed. The highest-value place to redeploy them is the bid.</description></item>
  <item><title>Data before models: why a uniform schema beats a smarter AI</title><link>https://fathygad.me/insights/data-before-models/</link><guid>https://fathygad.me/insights/data-before-models/</guid><pubDate>Sun, 25 Oct 2026 08:00:00 GMT</pubDate><description>Most AI disappointments in engineering are data problems. Before choosing a model, fix the library it reads from.</description></item>
  <item><title>Rules first. AI for judgment.</title><link>https://fathygad.me/insights/rules-first-ai-for-judgment/</link><guid>https://fathygad.me/insights/rules-first-ai-for-judgment/</guid><pubDate>Sun, 25 Oct 2026 08:00:00 GMT</pubDate><description>The fastest way to lose trust in AI is to use it for work that rules already do well. A practical split for engineering teams adopting AI.</description></item>
</channel>
</rss>
