<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>GD Data; Insights from the build</title><description>Practical insights on data engineering, AI in production, analytics, and cloud; written by the GD Data engineering team.</description><link>https://gddata.io/</link><language>en</language><item><title>Zero-downtime cloud migration: the phased playbook</title><link>https://gddata.io/blog/zero-downtime-migration/</link><guid isPermaLink="true">https://gddata.io/blog/zero-downtime-migration/</guid><description>Big-bang migrations are how projects fail. A phased, reversible approach keeps the business running while you modernise underneath it.</description><pubDate>Sat, 08 Aug 2026 00:00:00 GMT</pubDate><category>Cloud Migration</category><category>Architecture</category></item><item><title>From demo to production: a practical guide to RAG systems</title><link>https://gddata.io/blog/rag-in-production/</link><guid isPermaLink="true">https://gddata.io/blog/rag-in-production/</guid><description>A retrieval-augmented generation demo takes an afternoon; a reliable production RAG system takes real engineering. Here is what separates the two.</description><pubDate>Wed, 29 Jul 2026 00:00:00 GMT</pubDate><category>AI &amp; Machine Learning</category><category>LLM</category></item><item><title>Why your data pipelines keep breaking (and how data contracts fix it)</title><link>https://gddata.io/blog/data-contracts/</link><guid isPermaLink="true">https://gddata.io/blog/data-contracts/</guid><description>Silent schema changes are the number one cause of broken pipelines. Data contracts turn those failures into caught, owned, and versioned changes; here is how to adopt them.</description><pubDate>Wed, 15 Jul 2026 00:00:00 GMT</pubDate><category>Data Engineering</category><category>Reliability</category></item></channel></rss>