Over the last decade, hydrogen technologies have emerged as a key pillar in the transition toward clean and sustainable energy systems. Materials science underpins much of this progress, yet researchers face significant challenges in handling the vast and diverse data types involved – ranging from experimental characterizations and simulation outputs to industrial production parameters. Traditional data management solutions often struggle to maintain interoperability, consistency, and FAIR (Findable, Accessible, Interoperable, Reusable) compliance across these varied information sources. AIMWORKS addresses this complexity by developing an AI-driven knowledge graph (KG) framework specifically tailored to hydrogen-related materials research. In its agentic workflow, specialized AI agents coordinate tasks such as data gathering, preprocessing, and semantic mapping, turning heterogeneous inputs – like PDFs, spreadsheets, or imaging data – into an evolving, ontology-based knowledge graph. Leveraging Graph Retrieval Augmented Generation (GraphRAG), AIMWORKS can dynamically retrieve context from large datasets, enabling advanced semantic searches and more precise data queries. By using the Elementary Multiperspective Material Ontology (EMMO) and the Neo4j graph database, AIMWORKS ensures that metadata and relationships among materials, processes, and device applications are accurately represented and remain adaptable. This approach not only supports robust data interoperability across disciplines but also aligns with existing Helmholtz initiatives like unHIDE and Helmholtz-KG . Ultimately, AIMWORKS’ scalable framework empowers researchers and industry partners to accelerate discovery and innovation in sustainable hydrogen technologies, paving the way for more efficient and collaborative data-driven research. The goal of our project is to interlink information, make it explorable, and even discover knowledge that was previously considered in a different context or research field in order to generate new insights. The outcome of the project is a distributed knowledge graph built in a user-centered co-design process involving developers and use case partners from laboratories and large-scale facilities and various research domains.