This case study examines the development of a proof-of-concept (PoC) generative artificial intelligence (genAI) model inspired by OpenAI's ChatGPT®, implemented within the Office of Research Administration (ORA) at Emory University. Generative artificial intelligence (genAI) refers to AI models capable of producing human-like text. Specifically, this study shares practical insights and experiences from developing a private AI model tailored to support research administration operations. The initiative involved forming a specialized team, identifying a secure and efficient platform, and creating ORAgpt--a chatbot designed to provide Emory's research administrators with instant and accurate guidance on institutional policies, procedures, and administrative tasks. Key elements discussed include strategies for team selection, data curation, and model architecture, such as leveraging internal subject matter expertise, curating institutional documentation, and deploying cloud-based technology through Microsoft Azure. Despite challenges such as constrained funding and a compressed timeline, the project reached significant milestones, including positive stakeholder feedback and evidence indicating the model's potential to streamline tasks and enhance productivity. The findings underscore genAI's promise to transform research administration by increasing efficiency and providing a scalable framework for adoption by similar institutions. Future phases will incorporate structured governance, rigorous document vetting processes, and comprehensive financial planning to ensure sustainability.