<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>RedFin AI Blog</title>
    <link>https://www.redfinai.co.uk/blog.html</link>
    <atom:link href="https://www.redfinai.co.uk/rss.xml" rel="self" type="application/rss+xml" />
    <description>AI engineering, LLM cost optimisation, mobile development, MVP delivery and AI regulation — written by the RedFin AI engineering team.</description>
    <language>en-gb</language>
    <lastBuildDate>Wed, 15 Jul 2026 09:00:00 +0000</lastBuildDate>
    <generator>RedFin AI</generator>

    <item>
      <title>RAG Isn't a Product. Evaluation Is.</title>
      <link>https://www.redfinai.co.uk/blog/rag-evaluation-not-a-product.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/rag-evaluation-not-a-product.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>AI engineering</category>
      <description>Nine in ten Retrieval-Augmented Generation projects fail at evaluation, not retrieval. A practical framework for measuring RAG quality on your own data.</description>
    </item>

    <item>
      <title>Anatomy of a 4-Week MVP</title>
      <link>https://www.redfinai.co.uk/blog/anatomy-of-a-4-week-mvp.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/anatomy-of-a-4-week-mvp.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>Delivery</category>
      <description>Sprint by sprint, hour by hour: how RedFin AI compresses three months of typical agency work into 28 calendar days.</description>
    </item>

    <item>
      <title>How to Cut Your LLM Costs by 70%</title>
      <link>https://www.redfinai.co.uk/blog/cut-llm-costs-70-percent.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/cut-llm-costs-70-percent.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>Cost engineering</category>
      <description>Prompt caching, small-model routing, semantic caching, prompt compression and batching — the five production techniques we use to cut LLM bills.</description>
    </item>

    <item>
      <title>Native vs React Native in 2026</title>
      <link>https://www.redfinai.co.uk/blog/native-vs-react-native-2026.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/native-vs-react-native-2026.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>Mobile</category>
      <description>The 2018 answer doesn't apply anymore. Where the honest trade-offs actually are in 2026 for native vs cross-platform mobile development.</description>
    </item>

    <item>
      <title>The EU AI Act, Practically</title>
      <link>https://www.redfinai.co.uk/blog/eu-ai-act-practical-guide.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/eu-ai-act-practical-guide.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>AI regulation</category>
      <description>A practical guide to the EU AI Act for engineering and product teams. What applies to you, how to categorise your system, and paperwork that shortens sales cycles.</description>
    </item>

    <item>
      <title>Why "Prompt Engineering" Isn't Engineering — Yet</title>
      <link>https://www.redfinai.co.uk/blog/prompt-engineering-isnt-engineering.html</link>
      <guid isPermaLink="true">https://www.redfinai.co.uk/blog/prompt-engineering-isnt-engineering.html</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 +0000</pubDate>
      <category>AI engineering</category>
      <description>Most "prompt engineering" today involves no measurement, no versioning and no repeatability. Here's what the upgrade to real LLM engineering actually looks like.</description>
    </item>
  </channel>
</rss>
