<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Svolta · Insights</title><description>Engineering writing from Svolta on agents, quality checks, systems, and practical AI builds for operations teams.</description><link>https://svoltagroup.com/</link><language>en-au</language><item><title>AI agent infrastructure: the layer that makes agents safe to run</title><link>https://svoltagroup.com/insights/ai-agent-infrastructure/</link><guid isPermaLink="true">https://svoltagroup.com/insights/ai-agent-infrastructure/</guid><description>Agents fail in production for infrastructure reasons, not model reasons: no grounding, no evals, no audit trail, no data layer. The five components underneath every agent system that survives contact with a real queue.</description><pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate><category>Infrastructure</category><category>ai infrastructure</category><category>agents</category><category>production ai</category></item><item><title>Custom AI agent development services: what &apos;custom&apos; should mean</title><link>https://svoltagroup.com/insights/custom-ai-agent-development-services/</link><guid isPermaLink="true">https://svoltagroup.com/insights/custom-ai-agent-development-services/</guid><description>Every vendor sells &apos;custom&apos; AI agents. The word should mean your workflow, your systems, your data, your evals, and your ownership at handover. A checklist for reading proposals, and the parts that should never be custom.</description><pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate><category>Agents</category><category>ai agents</category><category>custom development</category><category>buying advice</category></item><item><title>Evals before agents: the order that saves the project</title><link>https://svoltagroup.com/insights/evals-before-agents/</link><guid isPermaLink="true">https://svoltagroup.com/insights/evals-before-agents/</guid><description>Write the eval set on your own data before you pick the framework, the model, or the vector database. Skip this and the project burns a quarter relitigating quality questions you could have answered in a week.</description><pubDate>Thu, 04 Jun 2026 00:00:00 GMT</pubDate><category>Evals</category><category>evals</category><category>agents</category><category>production</category><category>governance</category></item><item><title>Agentic workflows that survive production</title><link>https://svoltagroup.com/insights/agentic-workflows-production/</link><guid isPermaLink="true">https://svoltagroup.com/insights/agentic-workflows-production/</guid><description>A demo agent and a production agent share a vocabulary. They do not share an infrastructure. Caching, deterministic fallbacks, observability, eval gates on every change. The boring discipline most teams skip.</description><pubDate>Thu, 21 May 2026 00:00:00 GMT</pubDate><category>Agents</category><category>agents</category><category>production</category><category>infrastructure</category><category>reliability</category></item><item><title>Why your knowledge base isn&apos;t ready for RAG</title><link>https://svoltagroup.com/insights/knowledge-base-not-ready-for-rag/</link><guid isPermaLink="true">https://svoltagroup.com/insights/knowledge-base-not-ready-for-rag/</guid><description>Retrieval quality is a data hygiene problem before it is a vector-database problem. The corpus audit nobody runs is the work that decides whether the RAG pilot survives the third demo.</description><pubDate>Fri, 08 May 2026 00:00:00 GMT</pubDate><category>Retrieval</category><category>retrieval</category><category>rag</category><category>data</category><category>knowledge</category></item><item><title>Process mapping is the first AI artifact</title><link>https://svoltagroup.com/insights/process-mapping-as-the-first-artifact/</link><guid isPermaLink="true">https://svoltagroup.com/insights/process-mapping-as-the-first-artifact/</guid><description>Map the process before you write a prompt. The map is where you find the work an agent can actually do, and the implicit context that will kill the project if it stays hidden until integration.</description><pubDate>Fri, 03 Apr 2026 00:00:00 GMT</pubDate><category>Process</category><category>process</category><category>discovery</category><category>operations</category><category>diagnostics</category></item><item><title>Production AI infrastructure decisions for 2026</title><link>https://svoltagroup.com/insights/infrastructure-decisions-2026/</link><guid isPermaLink="true">https://svoltagroup.com/insights/infrastructure-decisions-2026/</guid><description>The decisions that matter in a mid-market AI build this year: model routing, retrieval over fine-tuning, hybrid search, caching, observability. And the ones the discourse is wrong about.</description><pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate><category>Infrastructure</category><category>infrastructure</category><category>architecture</category><category>models</category><category>observability</category></item></channel></rss>