<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel>
  <atom:link href="https://albearhalim.com/notes/feed.xml" rel="self" type="application/rss+xml"/>
  <title>Notes by Albear Halim</title><link>https://albearhalim.com/notes</link><description>Short essays on designing AI inside complex products.</description><language>en</language><lastBuildDate>Thu, 01 Oct 2026 00:00:00 +0000</lastBuildDate>
  <item><title>Design the failure first</title><link>https://albearhalim.com/notes/design-the-failure-first</link><guid>https://albearhalim.com/notes/design-the-failure-first</guid><pubDate>Thu, 01 Oct 2026 00:00:00 +0000</pubDate><description>What an AI answer should do when it times out, gets it wrong or runs out of budget, designed before the happy path.</description></item>
  <item><title>One row per proposed change</title><link>https://albearhalim.com/notes/one-row-per-proposed-change</link><guid>https://albearhalim.com/notes/one-row-per-proposed-change</guid><pubDate>Thu, 10 Sep 2026 00:00:00 +0000</pubDate><description>Why reviewing AI suggestions fails when the screen mirrors the database, and the one-sentence reframe that fixed it.</description></item>
  <item><title>Prototype inside the real product</title><link>https://albearhalim.com/notes/prototype-in-the-real-shell</link><guid>https://albearhalim.com/notes/prototype-in-the-real-shell</guid><pubDate>Thu, 20 Aug 2026 00:00:00 +0000</pubDate><description>Why I put low-fi screens inside a copy of the live product, at real scale, so stakeholders decide without imagining anything.</description></item>
</channel></rss>
