Recently, our society has experienced a huge transformative mark that has reshaped andreformed everything from organizational operations to the global economic systems. Thismark is called Generative AI. This thesis explores the development of an in-house AI chatbotin a Swedish public sector organization, which is intended to work as an automating andadministrative tool to help optimize service desk tasks and processes across departments.Through a case study approach and interviews of semi-structured nature with employees ata public sector organization in Sweden, tensions are being found in several areas of the AIplatform development processes. Thematic analysis is made of the collected data to delvedeeper into recurring trends among the respondents’ answers, highlighting the tensions theyare experiencing in developing and creating a Large Language Model Chatbot, named‘Digge’. This study defines the understanding of AI platform development and whatchallenges may arise, and further contributes to the narrow existing literature on AIplatform evolution.