{"id":25,"date":"2026-02-19T20:52:06","date_gmt":"2026-02-19T19:52:06","guid":{"rendered":"https:\/\/homeserver.meretsu.com\/?p=25"},"modified":"2026-02-19T20:52:06","modified_gmt":"2026-02-19T19:52:06","slug":"ki-projekt-am-forschungsinstitut-intelligente-dokumentenanalyse","status":"publish","type":"post","link":"https:\/\/homeserver.meretsu.com\/?p=25","title":{"rendered":"KI-Projekt am Forschungsinstitut: Intelligente Dokumentenanalyse&#8220;"},"content":{"rendered":"\n<p><strong>Kategorie:<\/strong> KI-Projekte<br><strong>Tags:<\/strong> LLM, Dokumentenanalyse, Datenschutz, Forschung<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Inhalt:<\/h3>\n\n\n\n<p>Unser Forschungsinstitut verarbeitet t\u00e4glich hunderte Dokumente \u2014 von Forschungsberichten bis zu Projektantr\u00e4gen. Die Frage war: K\u00f6nnen lokale LLMs diese Arbeit unterst\u00fctzen, ohne dass Daten das Haus verlassen?<\/p>\n\n\n\n<h4 class=\"wp-block-heading\">Das Problem<\/h4>\n\n\n\n<p>Kommerzielle KI-Dienste wie ChatGPT oder Claude API kommen f\u00fcr sensible Forschungsdaten nicht in Frage. DSGVO-Konformit\u00e4t und institutsinterne Richtlinien erfordern, dass alle Daten on-premise bleiben.<\/p>\n\n\n\n<h4 class=\"wp-block-heading\">Unsere L\u00f6sung<\/h4>\n\n\n\n<p>Ein lokaler AI-Server (siehe vorherigen Post) mit einer Pipeline aus:<\/p>\n\n\n\n<ol class=\"wp-block-list\">\n<li>Dokument-Upload \u00fcber internes Portal<\/li>\n\n\n\n<li>Text-Extraktion (OCR f\u00fcr gescannte Dokumente)<\/li>\n\n\n\n<li>Zusammenfassung und Schl\u00fcsselwort-Extraktion via LLM<\/li>\n\n\n\n<li>Ergebnisse in interner Datenbank<\/li>\n<\/ol>\n\n\n\n<h4 class=\"wp-block-heading\">Lessons Learned<\/h4>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Kleinere, spezialisierte Modelle schlagen oft die gro\u00dfen General-Purpose-Modelle<\/li>\n\n\n\n<li>Prompt Engineering macht 80% des Unterschieds<\/li>\n\n\n\n<li>Ein guter Fallback-Mechanismus ist Pflicht<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Kategorie: KI-ProjekteTags: LLM, Dokumentenanalyse, Datenschutz, Forschung Inhalt: Unser Forschungsinstitut verarbeitet t\u00e4glich hunderte Dokumente \u2014 von Forschungsberichten bis zu Projektantr\u00e4gen. Die Frage war: K\u00f6nnen lokale LLMs diese Arbeit unterst\u00fctzen, ohne dass&#8230;<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-25","post","type-post","status-publish","format-standard","hentry","category-allgemein"],"_links":{"self":[{"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/posts\/25","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=25"}],"version-history":[{"count":1,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/posts\/25\/revisions"}],"predecessor-version":[{"id":26,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=\/wp\/v2\/posts\/25\/revisions\/26"}],"wp:attachment":[{"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=25"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=25"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/homeserver.meretsu.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=25"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}