<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"
        xmlns:news="http://www.google.com/schemas/sitemap-news/0.9">
  <url>
    <loc>https://samtaelabs.com/blog/llm-adoption-decision-vs-alternatives</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-11T22:30:03.349Z</news:publication_date>
      <news:title>LLM이 꼭 필요한 기술처럼 느껴질 때, 그 느낌을 검증하는 방법</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/ai-agent-token-budget-orchestration-requirements-phase</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-11T12:30:36.461Z</news:publication_date>
      <news:title>AI 에이전트 대규모 투입 전, 요구사항 단계에서 정의하지 않으면 반드시 터지는 세 가지 비용 구조</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/open-source-ai-training-risk-fork-decision-guide</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-11T03:00:28.333Z</news:publication_date>
      <news:title>AI 학습에 쓰인 오픈소스를 채택할 때, 요구사항 단계에서 확인해야 할 세 가지 의사결정 지점</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/ai-agent-team-model-engineering-process-design</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-10T22:30:03.080Z</news:publication_date>
      <news:title>AI 에이전트를 팀으로 운영할 때, 복잡한 엔지니어링 작업이 반복 가능한 프로세스가 되는 조건</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/open-source-maintainer-risk-assessment-blueprint</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-10T12:30:36.257Z</news:publication_date>
      <news:title>오픈소스 의존성 진단: 유지보수자가 1~2명인 프로젝트를 보안 계획에 포함해야 하는 이유</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/game-benchmark-vs-real-decision-ai-vendor-selection</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-10-10T03:00:28.003Z</news:publication_date>
      <news:title>게임 점수가 높은 AI 에이전트, 실무 의사결정에서도 믿을 수 있는가</news:title>
    </news:news>
  </url>
</urlset>