An uncertain elite: Automation and rationalization of data science work in Germany

Addressing concerns about the displacement of high-skill labor, the presentation examines the automation potential in data engineering, data science, and data analytics, fields integral to the development and implementation of AI.

Using semi-structured interviews with professionals in German companies, the presentation reconstructs workflows and explores the interplay between task standardization and organizational dynamics—problem-solving, communication, and knowledge evolution—that limit automation feasibility. Data engineers’ tasks, like building data pipelines, resist automation due to the job requirements of problem-solving and collaboration. Similarly, data scientists’ model-building activities involve complex, evolving requirements. In contrast, data analysts focus on data visualization and face higher automation feasibility given standardized workflows.

The study underscores that, while automation enhances efficiency, it does not render high-skill work obsolete. These insights inform debates about employment and the evolving roles of data professionals in digital capitalism. The presentation ends with findings about the data scientists' perceptions of the role of employee representation.