<?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/backup-recovery-objectives-and-restore-testing</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-17T12:30:36.338Z</news:publication_date>
      <news:title>백업이 장애 때 실패하는 이유: 데이터별 복구 목표와 복구 훈련을 설계하는 법</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/healthcare-app-privacy-security-requirements-vendor-guide</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-17T03:00:28.285Z</news:publication_date>
      <news:title>의료 앱 개발 업체, 개인정보 보안 요구사항으로 고르는 법</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/sentry-grafana-observability-data-design-blueprint</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-16T22:30:03.245Z</news:publication_date>
      <news:title>Sentry와 Grafana 관측 데이터 설계: 사용자 식별은 줄이고 장애 원인은 남기는 운영 청사진</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/open-weight-llm-selection-failure-signals-and-validation</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-16T12:30:36.259Z</news:publication_date>
      <news:title>오픈 웨이트 LLM 도입이 실패하는 지점: 모델 점수보다 서비스 조건을 먼저 고정해야 하는 이유</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/local-llm-hardware-vs-api-total-cost-decision</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-16T03:00:28.610Z</news:publication_date>
      <news:title>로컬 LLM 장비 구매, API 청구서가 아니라 가동률과 업무 손실로 판단하는 법</news:title>
    </news:news>
  </url>
  <url>
    <loc>https://samtaelabs.com/blog/llm-judge-consensus-auto-approval-governance</loc>
    <news:news>
      <news:publication>
        <news:name>삼태연구소</news:name>
        <news:language>ko</news:language>
      </news:publication>
      <news:publication_date>2026-09-15T22:30:03.227Z</news:publication_date>
      <news:title>LLM 평가자 합의율로 자동 승인해도 될까: 공유 오류를 막는 운영 기준</news:title>
    </news:news>
  </url>
</urlset>