At the German Federal Institute for Risk Assessment (Bundesinstitut für Risikobewertung, BfR), Research Data Management (RDM) is defined as a component of the scientific research cycle. Therefore, every RDM aspect must be easy to integrate into the daily research routine. In combination with the application of digital methods and tools, a digital research paradigm emerges, which in turn is contextualized in various regulations, e.g. FAIR (Wilkinson et al. 2016), and dependencies (Hey et al. 2009). This in turn means that RDM concepts, methods and services need to be addressed on all levels of a public institution; political, strategic and operational levels for different stakeholders in research, IT, Open Science and management level. We therefore build our RDM on three pillars, in short: Strategy, Communication and Piloting. On the strategic and political level, the first pillar Strategy and Structure starts with the evaluation and use of strategy tools as well as awareness measures for management levels. This level needs to create and establish regulations and contexts that enable researchers to address RDM for their specific requirements (management as a service). The second pillar Communication and Teaching with the main components workshops, intranet and consulting is service-oriented for researchers. On the one hand, this pillar enables the researchers more and more to independently plan and implement RDM and, on the other hand, provides new impulses for further developments and improvements through feedback, discussions, etc. The third pillar Piloting and Establishing is dedicated to concrete implementation projects, by means of which RDM model implementations are developed together with the researchers, or/and departments at BfR, ideally with reference to all stations of the research data life cycle (vgl. Dierkes 2021). The data life cycle in turn is our guiding leitmotiv in our modular research-oriented RDM framework. The three pillars are the corner stones of our modular research-oriented RDM framework, which integrates in addition several aspects of agile and project management tools (e.g. agile requirements engineering, feedback and interfaces or traditional Gantt charts for planning and monitoring). These aspects and methods help to define ways for implementing RDM concepts and provide also a framework for assessing project risks as well as planning of human resources and budget planning. In addition, our framework will be further developed and established in close cooperation with IT infrastructure, research strategy, data privacy and quality management. This will provide a solid foundation for implementing RDM not as an isolated silo but as an accessible and operable structure at BfR. Our RDM framework allows to start processes on all levels in parallel and provides explicit interfaces between the pillars on the task level. For example, the research data policy is a strategic tool which is handed over in consulting tasks and is laying a foundation for pilot projects with researchers and with IT infrastructure. Starting in spring 2022, BfR has set up a pilot phase for implementing RDM. Our poster contribution provides deeper insights in our RDM framework and provides first examples of the implementation. References Dierkes, J. (2021) 4.1 Planung, Beschreibung und Dokumentation von Forschungsdaten. Praxishandbuch Forschungsdatenmanagement, eds. Putnings, Heike Neuroth und Janna Neumann, Berlin, Boston: De Gruyter Saur, pp. 303-326. doi: https://doi.org/10.1515/9783110657807-018. Hey, T., Tansley, S., Tolle, K. ed. (2009) The Fourth Paradigm: Data-intensive Scientific Discovery. Redmond WA: Microsoft Research. Wilkinson, M., Dumontier, M., Aalbersberg, I. et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data. 3, 160018. doi: https://doi.org/10.1038/sdata.2016.18.