<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Progressive on AI Engineering and Architecture</title><link>https://valery.tech/engineering/progressive/</link><description>Recent content in Progressive on AI Engineering and Architecture</description><generator>Hugo</generator><language>en-US</language><copyright>Copyright (c) 2014-2023</copyright><atom:link href="https://valery.tech/engineering/progressive/index.xml" rel="self" type="application/rss+xml"/><item><title>Backward Design Der</title><link>https://valery.tech/engineering/progressive/backward-design-der/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://valery.tech/engineering/progressive/backward-design-der/</guid><description>&lt;p&gt;What you were doing was closer to &lt;strong&gt;backward design derivation&lt;/strong&gt;:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;user need -&amp;gt; required outcome -&amp;gt; user workflow -&amp;gt; what the system must make possible at each step -&amp;gt; system responsibilities -&amp;gt; design&lt;/p&gt;</description></item><item><title>In Intent Driven</title><link>https://valery.tech/engineering/progressive/in-intent-driven/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://valery.tech/engineering/progressive/in-intent-driven/</guid><description>&lt;p&gt;In the context of software engineering and architecture, let&amp;rsquo;s explore approaches where user needs, Jobs-to-be-Done (JTBD), or other product-framing things serve as the foundational framework for reasoning about a system. I raise this because many traditional frameworks&amp;ndash;like the C4 model&amp;ndash;rely heavily on structural diagrams but often lack an underlying setting (forces, factors, context, etc.). They map out the architecture but fail to articulate &lt;em&gt;why&lt;/em&gt; a component exists in the first place. Beyond superficial interactions (e.g., &amp;rsquo;the user sends an email, and the system responds&amp;rsquo;), we need models that clarify the actual problem we are solving.&lt;/p&gt;</description></item></channel></rss>