Veranstaltungsprogramm

Traditional software development (TSD) relied on written specifications and detailed documentation (Mahboob et al., 2024) that caused excessive amounts of extra-work and, thus, slowed progress and often became obsolete when requirements changed during the project (Nanduri, 2025). That is why during the last two decades, the software industry more and more adopts Agile Software Development (ASD) approaches, such as Scrum and Kanban, for running flexibly through the project and focusing on the main task of software development (Tasneem et al., 2025; Oroy & John, 2024). ASD is guided by the “Agile Manifesto” (Beck et al., 2001), particularly through its second core value, which emphasizes “working software over comprehensive documentation”.

In today’s ASD, however, documentation remains essential, particularly in software development projects involving multiple developers and different stakeholders (Williams, 2024). In ASD practice, effective collaboration depends on documentation that serves as external team memory and enables knowledge transfer across the various roles and stakeholders in ASD teams (ASDT) (Brown et al., 2024). This, in turn, fosters informed participation and co-creation (Ates et al., 2025; Rosenbichler et al., 2020).

This is what can be referred to as the “documentation dilemma” in ASD (Nair, 2026; Canedo et al., 2024): Although documentation may limit flexibility, require additional effort, and hinder progress, it remains essential in ASD.

In our research we aim at understanding the documentation dilemma in ASD in detail. We particularly seek to analyze at which points in the ASD process documentation is executed by the developers and for what reasons. This provides us with a basis for the second step of our research, in which we aim to suggest the use of generative Artificial Intelligence (GenAI) with its ability to compose human-like texts as a “Cybernetic Teammate” that supports the documentation work in ASD teams (Dell'Acqua et al., 2025). Specifically, we develop a theoretical Human-in-the-Loop concept that illustrates a collaboration between this “Cybernetic Teammate” and the software developers. Our research is guided by the following research question:

How can GenAI support agile documentation to mitigate the documentation dilemma and facilitate shared understanding as a foundation for digital participation?