Researchers across scientific domains struggle with fragmented, incompatible data repositories and disconnected information silos. The project addresses the need for trustworthy AI systems that maintain human oversight at every critical decision point, aligned with GDPR and the forthcoming EU AI Act.

Building a trustworthy, FAIR-by-design agentic GenAI research assistant that helps scientists discover data, generate hypotheses, orchestrate workflows and produce reproducible research outputs inside the EOSC ecosystem.

OFFIS participates in the application domain energy systems, contributing its expertise in energy informatics to the implementation of the AI research assistant in this use case.

Agentic generative AI, FAIR Digital Objects, RO-Crates, cognitive Knowledge Graph technologies and human-in-the-loop oversight mechanisms; validated through three real-world domains: biodiversity, energy systems and materials science.