NFDI4ING’s Task Area on Assistance Systems and Artificial Intelligence presented its work at this year’s conference of the NFDI community for Data Science and Artificial Intelligence (NFDI4DataScience), taking place November 5-6 at the Weizenbaum Institute in Berlin. The conference was dedicated to the intersection between research data management (RDM) and artificial intelligence (AI). The overarching conversation revolved around two interrelated questions: how the data science and AI community can support the establishment of frameworks for fair and sustainable management of data, models and code on the one hand, and how AI can be used to lower the threshold for implementing fair RDM practices on the other.
The NFDI4DataScience conference offered an opportunity for NFDI4ING to showcase its work in the field of AI and RDM. Over the past year and a half, a dedicated working group in NFDI4ING – the Task Area on Assistance Systems and Artificial Intelligence – has been developing AI-assisted tools to help engineering researchers implement RDM practices early on. Development takes place along two main veins: an independent web-based service and integrated conversational assistants.
Titled NFDI4ING AI Assistant, the independent web-based service, which is currently under development, consolidates existing tools, RDM guidelines for engineering disciplines, and research output from NFDI4ING. A question-answering agent guides researchers through the landscape of tools and services for RDM in engineering: from finding datasets and code, metadata standards and terminologies, to education and training opportunities on best RDM practices for the engineering community. Beyond question-answering, the NFDI4ING AI Assistant is planned to grow into a single point of access to the tools of the Consortium: using Model Context Protocol, it should directly connect users to NFDI4ING services, such as the Research Software Finder. Last but not least, the AI Assistant’s knowledge base is accessible through an API to external services, paving the way for integration across the consortium.
In the scope of this working group, two conversational assistants have been integrated in existing NFDI4ING services: the Research Data Management Organizer (RDMO NFDI4ING) and the NFDI4ING Jupyter Service. The two assistants support researchers while filling out a data management plan in RDMO NFDI4ING, or with coding tasks and CodeMeta files in NFDI4ING Jupyter Service.
The NFDI4DataScience 2026 Conference was an excellent opportunity to both share our experience and to learn from ongoing work in AI and RDM in other NFDI consortia and higher education institutions. Combined with the insights shared by the Data Science and Artificial Intelligence community, this experience will help us shape the next phase in the development of AI-supported RDM services.
Elena Volkanovska