<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>AppRiddle — Applied engineering research</title>
    <link>https://appriddle.com/insights</link>
    <description>AppRiddle builds enterprise AI systems you can measure, and the platform and cloud engineering that keeps them running. Evaluated from the first commit.</description>
    <language>en</language>
    <lastBuildDate>Sun, 30 Aug 2026 00:00:00 GMT</lastBuildDate>
    <atom:link href="https://appriddle.com/rss.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>Pay for an agent only where steps are unpredictable</title>
      <link>https://appriddle.com/insights/pay-for-an-agent-where-steps-are-unpredictable</link>
      <guid isPermaLink="true">https://appriddle.com/insights/pay-for-an-agent-where-steps-are-unpredictable</guid>
      <pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate>
      <description>You are choosing between a system that decides what to look up once and one that decides on every step. At list prices, eight steps bill roughly ten times more.</description>
      <category>Agents</category>
      <category>RAG</category>
      <category>Cost</category>
      <category>Tool use</category>
      <category>Prompt injection</category>
    </item>
    <item>
      <title>Microservices: when the boundary earns its cost, and when it does not</title>
      <link>https://appriddle.com/insights/microservices-when-and-when-not</link>
      <guid isPermaLink="true">https://appriddle.com/insights/microservices-when-and-when-not</guid>
      <pubDate>Thu, 30 Jul 2026 00:00:00 GMT</pubDate>
      <description>Service count is not the goal — independent deployability is, and the two are frequently unrelated. A decision framework, the arithmetic a network boundary adds, and the five situations where the correct answer is one service.</description>
      <category>Architecture</category>
      <category>Microservices</category>
      <category>Modular monolith</category>
      <category>Team topology</category>
    </item>
    <item>
      <title>Cloud observability: instrument everything, keep almost none of it</title>
      <link>https://appriddle.com/insights/enterprise-cloud-observability</link>
      <guid isPermaLink="true">https://appriddle.com/insights/enterprise-cloud-observability</guid>
      <pubDate>Sun, 26 Jul 2026 00:00:00 GMT</pubDate>
      <description>The three questions an on-call engineer actually needs answered, the signals that answer them, and why your bill is set by cardinality and retention rather than traffic. With Collector config, burn-rate alerts, and list prices from six platforms.</description>
      <category>Observability</category>
      <category>OpenTelemetry</category>
      <category>Cost control</category>
      <category>SLOs</category>
    </item>
    <item>
      <title>Multi-tenant foundations: the work before the feature work</title>
      <link>https://appriddle.com/insights/multi-tenant-saas-foundations</link>
      <guid isPermaLink="true">https://appriddle.com/insights/multi-tenant-saas-foundations</guid>
      <pubDate>Tue, 14 Jul 2026 00:00:00 GMT</pubDate>
      <description>Tenancy model, row-level isolation, SSO and RBAC, metering, and an audit log that survives examination — with the migrations, the connection-pool trap that defeats RLS, and the test that tries to leak.</description>
      <category>Multi-tenancy</category>
      <category>PostgreSQL</category>
      <category>RBAC</category>
      <category>Audit</category>
    </item>
    <item>
      <title>A staged path to multi-region on AWS</title>
      <link>https://appriddle.com/insights/staged-path-to-multi-region-aws</link>
      <guid isPermaLink="true">https://appriddle.com/insights/staged-path-to-multi-region-aws</guid>
      <pubDate>Tue, 09 Jun 2026 00:00:00 GMT</pubDate>
      <description>Four stages, each removing one named failure mode, with the RTO and RPO each one actually buys. Most teams should stop at stage two — here is how to tell whether you are one of them.</description>
      <category>AWS</category>
      <category>Resilience</category>
      <category>Terraform</category>
      <category>Aurora</category>
    </item>
    <item>
      <title>Choosing a RAG architecture for enterprise knowledge</title>
      <link>https://appriddle.com/insights/rag-architecture-enterprise-knowledge</link>
      <guid isPermaLink="true">https://appriddle.com/insights/rag-architecture-enterprise-knowledge</guid>
      <pubDate>Tue, 12 May 2026 00:00:00 GMT</pubDate>
      <description>Chunking, hybrid retrieval, reranking, and the evaluation harness that decides whether any of it worked. With the pgvector schema and the CI gate we actually ship.</description>
      <category>RAG</category>
      <category>pgvector</category>
      <category>Evaluation</category>
      <category>Retrieval</category>
    </item>
  </channel>
</rss>
