Database preservation frequently happens post-factum: databases are transferred and migrated into preservation formats and environments after a project has ended. This increases the risks concerning incompatibility and pushes the preservation burden after the initial lifetime and use of the data. We propose a database repository infrastructure, where databases are created, used and preserved directly in the data curation environment. This increases the FAIRness of the data curated as professional data stewardship activities accompany the databases right from the onset. We present the FAIR Data Austria Database Repository (FDA-DBRepo) infrastructure and provide a first version of an open-source reference implementation.
