<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Tech Company Blogs on MacWorks</title><link>https://macworks.dev/docs/month/tech/</link><description>Recent content in Tech Company Blogs on MacWorks</description><generator>Hugo</generator><language>en</language><atom:link href="https://macworks.dev/docs/month/tech/index.xml" rel="self" type="application/rss+xml"/><item><title>Week 13 Summary</title><link>https://macworks.dev/docs/month/tech/weekly-2026-W13/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://macworks.dev/docs/month/tech/weekly-2026-W13/</guid><description>&lt;h1 id="engineering--scale--week-of-2026-03-19-to-2026-03-26"&gt;Engineering @ Scale — Week of 2026-03-19 to 2026-03-26&lt;a class="anchor" href="#engineering--scale--week-of-2026-03-19-to-2026-03-26"&gt;#&lt;/a&gt;&lt;/h1&gt;
&lt;h2 id="week-in-review"&gt;Week in Review&lt;a class="anchor" href="#week-in-review"&gt;#&lt;/a&gt;&lt;/h2&gt;
&lt;p&gt;The industry is aggressively pivoting away from treating LLMs as magical black boxes, instead wrapping them in rigorous, deterministic software primitives. Across the stack, engineering organizations are moving to isolate non-deterministic AI tasks from synchronous user requests, heavily sandboxing agentic execution at the infrastructure level, and grappling with the reality that AI code generation is simply shifting the engineering bottleneck from typing to system specification and verification.&lt;/p&gt;</description></item><item><title>Week 14 Summary</title><link>https://macworks.dev/docs/month/tech/weekly-2026-W14/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://macworks.dev/docs/month/tech/weekly-2026-W14/</guid><description>&lt;h1 id="engineering--scale--week-of-2026-03-28-to-2026-04-03"&gt;Engineering @ Scale — Week of 2026-03-28 to 2026-04-03&lt;a class="anchor" href="#engineering--scale--week-of-2026-03-28-to-2026-04-03"&gt;#&lt;/a&gt;&lt;/h1&gt;
&lt;h2 id="week-in-review"&gt;Week in Review&lt;a class="anchor" href="#week-in-review"&gt;#&lt;/a&gt;&lt;/h2&gt;
&lt;p&gt;The industry is moving past the novelty of generative AI, focusing instead on bounding autonomous agents with strict architectural contracts, standardizing machine-to-machine context layers, and pushing security enforcement to the absolute edge. Concurrently, legacy infrastructure assumptions—ranging from traditional LRU caching algorithms to deeply nested UI component trees—are failing under the weight of AI-driven traffic and massive data scale, forcing engineers to adopt zero-trust capability sandboxing and highly optimized, O(1) data access patterns.&lt;/p&gt;</description></item></channel></rss>