As media archives begin to adopt AI-driven metadata workflows, a strategic question arises: should they hire their own AI specialists, Data Scientists and Analysts or rely on central IT resources or external support? This presentation argues that the challenge is no longer primarily technical, but organizational. Successful implementation requires new roles, workflows, and competencies—making the case for dedicated data profiles within the archive itself.
This is not just theory: At both the ZDF archive and RSI archive, first-hand experience is being gathered through pilot initiatives. At ZDF, a one-year position for Data Science/Analytics was created in order to develop data-driven services, detect patterns in metadata and co-design semantic models for media archives. The onboarding raised key questions around operational readiness, team integration, IT collaboration, and the scope of Data-related tasks and roles. These reflections are shaping not only the role itself but the surrounding organizational structures.
By the time of the vfm-Tagung, both institutions will be able to share early insights into what it means to embed such expertise directly within archival teams.