{"id":32543,"date":"2025-03-20T10:31:08","date_gmt":"2025-03-20T09:31:08","guid":{"rendered":"https:\/\/opi.org.pl\/?post_type=opi_project&#038;p=32543"},"modified":"2025-09-08T13:11:56","modified_gmt":"2025-09-08T11:11:56","slug":"hive-ai-rozwoj-i-pilotazowe-wdrozenie-duzych-modeli-jezykowych-w-polskiej-administracji-publicznej","status":"publish","type":"opi_project","link":"https:\/\/opi.org.pl\/en\/project\/hive-ai-rozwoj-i-pilotazowe-wdrozenie-duzych-modeli-jezykowych-w-polskiej-administracji-publicznej\/","title":{"rendered":"HIVE AI: Development and pilot implementation of large language models in Polish public administration"},"content":{"rendered":"<section class=\"project-data\">\r\n\t<h2>Project data<\/h2>\r\n\t<div class=\"opi-table-wrapper\"><table class=\"opi-project-details\">\r\n\t\t<tbody>\r\n<tr class=\"_project_date_start\"><td><strong>Project start date<\/strong><\/td><td class=\"date\">2025-02-20<\/td><\/tr><tr class=\"_project_date_end\"><td><strong>Project end date<\/strong><\/td><td class=\"date\">2025-12-31<\/td><\/tr><tr class=\"_project_cost\"><td><strong>Project cost<\/strong><\/td><td class=\"number\">18,983,055.00 PLN<\/td><\/tr>\t\t<\/tbody>\r\n\t<\/table><\/table>\r\n<\/section>\r\n\r\n<section class=\"project-partners\"><h3>Lider<\/h3><ul><li>Research and Academic Computer Network<\/li><\/ul><h3>Partners<\/h3><ul><li>AGH University of Krakow<\/li><li>Centre for Information Technology<\/li><li>Institute of Computer Science of the Polish Academy of Sciences<\/li><li>Institute of Slavic Studies of the Polish Academy of Sciences<\/li><li>National Information Processing Institute <\/li><li>Wroc\u0142aw University of Science and Technology<\/li><li>University of Lodz<\/li><\/ul><\/section>\n<p>HIVE AI is a direct continuation of the <em>Polish Large Language Model (PLLuM)<\/em> project. HIVE AI aims to facilitate the integration of Polish language models into the mObywatel app, develop a clerical assistant for the Polish Ministry of Digital Affairs, and conduct a pilot implementation of LLMs in a municipal or provincial office.<br>To implement LLMs and adapt them to specific needs of public administration identified during the pilot phase, new linguistic datasets will be collected. Additionally, new instruction fine-tuning datasets will be developed, comprising prompts (queries or commands) and expected responses. Depending on the type of instruction, datasets may include additional elements such as reasoning, context, and keywords. New datasets for aligning LLMs with human preferences will also be created, containing sets of prompts and multiple response variations that are assessed for quality.<\/p>\n\n\n\n<p>Based on these resources, the PLLuM model family will be expanded to comprise new models of various sizes, including:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>General-purpose models, which are designed for a wide range of linguistic tasks, with a particular focus on administrative environment<\/li>\n\n\n\n<li>Generators, which are specialised Retrieval-Augmented Generation (RAG) models whose objective is to generate final responses based on natural language queries and retrieved contextual documents that are most relevant to particular queries.<\/li>\n<\/ul>\n\n\n\n<p>The project will involve close collaboration with public institutions in which the pilot implementation will roll out, ensuring efficient support and automation of their operations.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Funding<\/strong><\/h3>\n\n\n\n<p>The project is funded by the Ministry of Digital Affairs as part of budget item 27 \u2018Informatisation in 2025\u2019.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"768\" height=\"253\" src=\"https:\/\/opi.org.pl\/wp-content\/uploads\/2025\/03\/minister-of-digital-affairs-republic-of-poland-logo.webp\" alt=\"\" class=\"wp-image-32546\" style=\"width:306px;height:auto\" srcset=\"https:\/\/opi.org.pl\/wp-content\/uploads\/2025\/03\/minister-of-digital-affairs-republic-of-poland-logo.webp 768w, https:\/\/opi.org.pl\/wp-content\/uploads\/2025\/03\/minister-of-digital-affairs-republic-of-poland-logo-300x99.webp 300w\" sizes=\"auto, (max-width: 768px) 100vw, 768px\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-full is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"700\" height=\"233\" src=\"https:\/\/opi.org.pl\/wp-content\/uploads\/2024\/02\/flaga-i-godlo.png\" alt=\"\" class=\"wp-image-28274\" style=\"width:304px;height:auto\" srcset=\"https:\/\/opi.org.pl\/wp-content\/uploads\/2024\/02\/flaga-i-godlo.png 700w, https:\/\/opi.org.pl\/wp-content\/uploads\/2024\/02\/flaga-i-godlo-300x100.png 300w\" sizes=\"auto, (max-width: 700px) 100vw, 700px\" \/><\/figure>\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>The goal of this project is to support the digitalisation and automation of routine administrative tasks in Poland through the pilot implementation of Polish large language models (LLMs) in public administration systems.<\/p>\n","protected":false},"featured_media":32547,"template":"","meta":{"_links_to":"","_links_to_target":""},"class_list":["post-32543","opi_project","type-opi_project","status-publish","has-post-thumbnail","hentry"],"_links":{"self":[{"href":"https:\/\/opi.org.pl\/en\/wp-json\/wp\/v2\/opi_project\/32543","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/opi.org.pl\/en\/wp-json\/wp\/v2\/opi_project"}],"about":[{"href":"https:\/\/opi.org.pl\/en\/wp-json\/wp\/v2\/types\/opi_project"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/opi.org.pl\/en\/wp-json\/wp\/v2\/media\/32547"}],"wp:attachment":[{"href":"https:\/\/opi.org.pl\/en\/wp-json\/wp\/v2\/media?parent=32543"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}