Research Paper Presented at DESRIST 2026 in Münster

At the 21st International Conference on Design Science Research in Information Systems and Technology (DESRIST 2026), Flora Horn presented a research paper in the field of Generative AI and software development.

Supporting Information Seeking in Software Development: A Design Theory for Local Retrieval-Augmented Generation Across Distributed Knowledge Sources
Flora Horn, Martin Wiener

The paper addresses the challenge that software developers spend a substantial portion of their working time searching for information across distributed knowledge sources such as wikis, issue tracking systems, and repositories. Using a Design Science Research approach, the study develops a design theory for local Retrieval-Augmented Generation (RAG) systems tailored to the needs of privacy-sensitive and resource-constrained organizations. The design theory was instantiated in a prototype and evaluated in a software company from the SME sector (small and medium-sized enterprises). The results demonstrate how local RAG systems can improve access to information across distributed knowledge sources while enabling the adoption of Generative AI in privacy-sensitive and resource-constrained environments. By processing data locally, sensitive organizational information remains within the company’s IT infrastructure and does not need to be transferred to external services.