Since 2019, dataNooS (http://datanoos.univ-toulouse.fr/en/) is an academic alliance bringing together members of the academic community in Toulouse area to deal with issues related to digital resource sharing and knowledge practices. Thanks to use cases, we collect, compare and share research experiences related to the production, management and valorization of research data. We carry out a transdisciplinary process in coherence with (inter)national and disciplinary initiatives. Thanks to several experimental projects, we are currently implementing a generic and integrated platform to manage cross-disciplinary datasets. It supports dataset description with semantic metadata and their query through a faceted search interface. A prototype of the platform provides several services such as:•setting up a domain ontology adapted from a generic one to select the vocabulary for metadata descriptions;•designing templates of metadata based on this ontology, so that the quality of the metadata be checked;•a faceted search device that guides the input of search criteria thanks to the ontology;•a dataset visualization and browsing service.•repositories of digital objects (datasets, queries, templates, workflows) and knowledge (ontologies, vocabularies, catalogs, services directories);This platform is connected to the heterogeneous Open science eco-systems. Search requests are based on shared or open data repositories, metadata repositories and data portals to reference disseminated data on one point.This platform also integrates users' experience and feedback to provide personalized assistance. The platform will also support how to use the datasets in workflows, and in data processing environments.The poster presents the FAIR principles and ontologies that guided the design of the Platform. It will expose its architecture and the modules already implemented in the prototype.