<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Middleware on Muhammad Hassan Raza</title><link>https://mhassan.dev/tags/middleware/</link><description>Recent content in Middleware on Muhammad Hassan Raza</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sat, 10 Jan 2026 10:00:00 +0500</lastBuildDate><atom:link href="https://mhassan.dev/tags/middleware/index.xml" rel="self" type="application/rss+xml"/><item><title>Multi-Agent LLM Middleware: Lessons from Marketing Accelerant</title><link>https://mhassan.dev/blog/langgraph-multi-agent-middleware/</link><pubDate>Sat, 10 Jan 2026 10:00:00 +0500</pubDate><guid>https://mhassan.dev/blog/langgraph-multi-agent-middleware/</guid><description>What it took to make a multi-agent AI stack survive production: model routing, approval gates, retries, cost control, and sane control flow.</description></item></channel></rss>