The sustainable handling of research data requires collaboration across stakeholders [1], giv-en the challenges arising from the development of workflows, standards, and infrastructures for effective research data management (RDM). Involved stakeholders include central units (e.g. libraries, IT) and institutions, offering infrastructures and services, and researchers with specific needs and expectations with regard to RDM services. The NFDI consortia and sections provide an important foundation for these challenges. How-ever, institutional engagement and support remain essential, and cross-institutional structures like the RDM federal state initiatives actively support this goal. In both Schleswig-Holstein and Mecklenburg-Vorpommern, the corresponding federal state initiatives FDM-SH [2,3] and the proposed concept FDM-MV [4] implement a community-based approach enabling stakeholders to collaboratively address challenges through interest and working groups. The added values of the community-based approach include: 1. demand-oriented: The developed solutions are consistently geared towards the needs of community members to prevent developments from being driven forward that do not meet the needs of local research institutions. 2. avoidance of isolated solutions: The joint and collaborative development of ideas and solutions avoids isolated individual measures, saves resources in the long term and strengthens supra-regional con-nectivity. 3. flexibility and dynamism: The interest and working groups as expression of the active community react flexibly to new developments in the field of RDM. They can be dynamically set up or paused as required – without the need for large amounts of resources. Current topics can thus be addressed quickly and worked on jointly. 4. sustainable knowledge transfer: The community-based approach promotes a reciprocal, lively exchange of knowledge between the participating organisations – a win-win situation for everyone involved. Since 2023, four interest and three working groups [5] have emerged within the active FDM-SH-community in Schleswig-Holstein. Topics range from those relevant to central units – such as storage, archiving, and policies – to those raised by researchers, including sustainable handling of research software and molecular biological data (especially OMICS). In coordina-tion with the NFDI and in exchange with representatives from Schleswig-Holstein, other federal states and with networks such as ELIXIR, the objectives are setting up RDM systems for sequence (meta)data storage and data brokering as well as adopting and developing interop-erable (meta)data standards. Five working groups were established within the "Datenkompass MV" project, which was a demonstrator for an RDM-community in Mecklenburg-Vorpommern. They address basic services and training, as well as structural aspects such as governance, legal frameworks, and establishing a task catalogue for the RDM network. Notably, the working group for Teaching and Training successfully brought together diverse stakeholders to map existing training programmes, identify gaps, and explore synergies – using the learning objectives matrix on RDM [6] as a foundation. The analysis revealed a broad spectrum of offerings, from interdisciplinary short courses to semester-long lectures and subject-specific modules. With our presentation, we showcase the added value of community-based approaches in Schleswig-Holstein and Mecklenburg-Vorpommern and underline their potential for collaboration with the NFDI. The state initiatives function as information hubs – communicating NFDI developments to institutions and target groups, and helping disseminate NFDI resources across regions.