<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Production on IQLAS</title>
    <link>/tags/production/</link>
    <description>Recent content in Production on IQLAS</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Sat, 11 Apr 2026 17:17:57 +0530</lastBuildDate>
    <atom:link href="/tags/production/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Serverless in Production: Real Patterns with AWS Lambda</title>
      <link>/blog/serverless-in-production-real-patterns-with-aws-lambda/</link>
      <pubDate>Wed, 18 Feb 2026 00:00:00 +0000</pubDate>
      <guid>/blog/serverless-in-production-real-patterns-with-aws-lambda/</guid>
      <description>Serverless is not magic — it is a different set of tradeoffs. Here are the patterns that work in production, the anti-patterns that bite you at scale, and the mental models that make Lambda make sense.</description>
    </item>
    <item>
      <title>Retrieval-Augmented Generation: Building LLMs That Know What They Don&#39;t Know</title>
      <link>/blog/retrieval-augmented-generation-building-llms-that-know-what-they-dont-know/</link>
      <pubDate>Thu, 15 Jan 2026 00:00:00 +0000</pubDate>
      <guid>/blog/retrieval-augmented-generation-building-llms-that-know-what-they-dont-know/</guid>
      <description>RAG is not a buzzword — it is a practical architecture that grounds language models in verified knowledge. Here is how it works, why the naive version fails, and what production RAG actually looks like.</description>
    </item>
    <item>
      <title>MLOps in Practice: Building Machine Learning Systems That Last</title>
      <link>/blog/mlops-in-practice-building-machine-learning-systems-that-last/</link>
      <pubDate>Tue, 28 Oct 2025 00:00:00 +0000</pubDate>
      <guid>/blog/mlops-in-practice-building-machine-learning-systems-that-last/</guid>
      <description>Getting a model to 90% accuracy is the easy part. Getting it to 90% accuracy six months later, in production, with real data and real consequences, is where MLOps begins.</description>
    </item>
  </channel>
</rss>
