<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Alien6 Research</title>
    <link>https://alien6.com/en/research/articles/</link>
    <atom:link href="https://alien6.com/en/research/feed.xml" rel="self" type="application/rss+xml"/>
    <description>Articles from the Alien6 Research Hub — AI, architecture, engineering and security.</description>
    <language>en</language>
    <lastBuildDate>Tue, 03 Nov 2026 08:00:00 GMT</lastBuildDate>
    <item>
      <title>AI in production: the pitfalls of LLM quota management</title>
      <link>https://alien6.com/en/research/articles/ai-in-production-llm-quota-management-pitfalls/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/ai-in-production-llm-quota-management-pitfalls/</guid>
      <description>An experimental study of shared quotas, queues and responses that reach the client too late.</description>
      <pubDate>Sun, 15 Nov 2026 08:00:00 GMT</pubDate>
      <category>LLM</category>
      <category>Kubernetes</category>
      <category>Architecture</category>
      <category>Production</category>
    </item>
    <item>
      <title>RAG: understanding the path from retrieval to response</title>
      <link>https://alien6.com/en/research/articles/rag-architecture-audit/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/rag-architecture-audit/</guid>
      <description>We followed RAG retrieval results through the text sent to the model and the evaluated answer. We want to identify which stage to change and how to measure its effects.</description>
      <pubDate>Sun, 15 Nov 2026 08:00:00 GMT</pubDate>
      <category>RAG</category>
      <category>Architecture</category>
      <category>Embeddings</category>
      <category>Production</category>
    </item>
    <item>
      <title>Give your agents an identity!</title>
      <link>https://alien6.com/en/research/articles/identite-autorite-agents/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/identite-autorite-agents/</guid>
      <description>Your agents can act on your behalf. Give them a verifiable identity, limited permissions and traceable actions, from assignment to execution.</description>
      <pubDate>Tue, 06 Oct 2026 08:00:00 GMT</pubDate>
      <category>Zero-Trust</category>
      <category>Sécurité</category>
      <category>Architecture</category>
      <category>LLM</category>
    </item>
    <item>
      <title>Semantic drift: the invisible risk in multi-agent systems</title>
      <link>https://alien6.com/en/research/articles/semantic-drift-multi-agents/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/semantic-drift-multi-agents/</guid>
      <description>When several agents share the same words but not quite the same business definitions, every component can work correctly while the system as a whole quietly starts to drift.</description>
      <pubDate>Tue, 15 Sep 2026 08:00:00 GMT</pubDate>
      <category>Agentic AI</category>
      <category>Architecture</category>
      <category>Semantic Layer</category>
      <category>Ontology</category>
    </item>
    <item>
      <title>SPIFFE: a stable identity for constantly evolving applications</title>
      <link>https://alien6.com/en/research/articles/spiffe-identite-workloads/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/spiffe-identite-workloads/</guid>
      <description>Giving applications and services a verifiable identity, independent of everything that keeps changing in their runtime environment — the second Zero Trust foundation of our projects.</description>
      <pubDate>Mon, 31 Aug 2026 08:00:00 GMT</pubDate>
      <category>Zero-Trust</category>
      <category>Sécurité</category>
      <category>Kubernetes</category>
    </item>
    <item>
      <title>eBPF: trust put to the test of the cloud</title>
      <link>https://alien6.com/en/research/articles/ebpf-confiance-cloud/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/ebpf-confiance-cloud/</guid>
      <description>At Alien6, we are experimenting with eBPF to make it one of the foundations of our application architecture: containing incidents, easing audits and preserving service continuity.</description>
      <pubDate>Wed, 26 Aug 2026 08:00:00 GMT</pubDate>
      <category>Zero-Trust</category>
      <category>Sécurité</category>
      <category>Infrastructure</category>
    </item>
    <item>
      <title>When the AI agent starts forgetting what it already knows</title>
      <link>https://alien6.com/en/research/articles/watchdog-gouvernance-agents-ia/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/watchdog-gouvernance-agents-ia/</guid>
      <description>With Watchdog, Alien6&apos;s R&amp;D explores how to measure, contain and improve the work of development agents without sacrificing quality or turning their governance into a surveillance system.</description>
      <pubDate>Tue, 04 Aug 2026 08:00:00 GMT</pubDate>
      <category>LLM</category>
      <category>Architecture</category>
      <category>Production</category>
    </item>
    <item>
      <title>Technical OSINT: Mapping Attack Surfaces</title>
      <link>https://alien6.com/en/research/articles/osint-cartographie-menaces/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/osint-cartographie-menaces/</guid>
      <description>Passive and active reconnaissance techniques for mapping your own attack surface and identifying exposure vectors.</description>
      <pubDate>Sun, 15 Mar 2026 08:00:00 GMT</pubDate>
      <category>Sécurité</category>
      <category>OSINT</category>
      <category>Threat Intelligence</category>
    </item>
    <item>
      <title>Cryptographic Attestation in CI/CD Pipelines</title>
      <link>https://alien6.com/en/research/articles/attestation-cryptographique-cicd/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/attestation-cryptographique-cicd/</guid>
      <description>How to use cryptographic attestation to prevent supply chain attacks in CI/CD pipelines.</description>
      <pubDate>Sun, 15 Feb 2026 08:00:00 GMT</pubDate>
      <category>DevSecOps</category>
      <category>Sécurité</category>
      <category>CI/CD</category>
    </item>
    <item>
      <title>Embedding Strategy for Advanced RAG Systems</title>
      <link>https://alien6.com/en/research/articles/embedding-strategy-rag/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/embedding-strategy-rag/</guid>
      <description>Design an embedding strategy around retrieval evidence: dense and sparse search, re-ranking, dated benchmarks and a reproducible rank-fusion example.</description>
      <pubDate>Thu, 12 Nov 2026 08:00:00 GMT</pubDate>
      <category>RAG</category>
      <category>Embeddings</category>
      <category>Architecture</category>
    </item>
    <item>
      <title>Why Your RAG Pipeline Is Underperforming</title>
      <link>https://alien6.com/en/research/articles/rag-pipeline-optimisation/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/rag-pipeline-optimisation/</guid>
      <description>The four structural pathologies that cause most production RAG systems to underperform — and how to fix them.</description>
      <pubDate>Mon, 15 Dec 2025 08:00:00 GMT</pubDate>
      <category>RAG</category>
      <category>LLM</category>
      <category>Architecture</category>
    </item>
    <item>
      <title>Prompt Engineering vs Fine-Tuning: Which Strategy to Choose</title>
      <link>https://alien6.com/en/research/articles/prompt-vs-finetuning/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/prompt-vs-finetuning/</guid>
      <description>Start with prompting, use RAG for changing source knowledge, and fine-tune only for a measured behavioural gap. A decision framework with dated evidence and reproducible evaluation.</description>
      <pubDate>Thu, 12 Nov 2026 08:00:00 GMT</pubDate>
      <category>LLM</category>
      <category>Prompt Engineering</category>
      <category>Fine-Tuning</category>
    </item>
    <item>
      <title>LLMs in Production: 5 Architecture Patterns</title>
      <link>https://alien6.com/en/research/articles/llm-production-patterns/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/llm-production-patterns/</guid>
      <description>Five patterns observed in production-grade LLM systems for managing cost, latency, and reliability.</description>
      <pubDate>Wed, 15 Oct 2025 08:00:00 GMT</pubDate>
      <category>LLM</category>
      <category>Architecture</category>
      <category>Production</category>
    </item>
    <item>
      <title>Zero-Trust Is Not a Product, It&apos;s a Philosophy</title>
      <link>https://alien6.com/en/research/articles/zero-trust-philosophie/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/zero-trust-philosophie/</guid>
      <description>Why zero-trust cannot be purchased as a product, and what it actually requires architecturally.</description>
      <pubDate>Mon, 15 Sep 2025 08:00:00 GMT</pubDate>
      <category>Sécurité</category>
      <category>Zero-Trust</category>
      <category>Architecture</category>
    </item>
    <item>
      <title>The Kubernetes Operator as a General Pattern</title>
      <link>https://alien6.com/en/research/articles/kubernetes-operateur-pattern/</link>
      <guid isPermaLink="true">https://alien6.com/en/research/articles/kubernetes-operateur-pattern/</guid>
      <description>How the Kubernetes operator pattern generalizes beyond stateful apps to encode any complex operational logic into the control plane.</description>
      <pubDate>Fri, 15 Aug 2025 08:00:00 GMT</pubDate>
      <category>Infrastructure</category>
      <category>Kubernetes</category>
      <category>DevOps</category>
    </item>
  </channel>
</rss>
