Project plan

Pillar 1 – Governance, policy and EOSC alignment (WP1–WP6). The first pillar establishes the governance, ethical, and technical foundations of the project.

• WP1 and WP2 provide overall project management, scientific coordination, and quality assurance, ensuring that progress remains coherent and risks are managed proactively.
• WP3 sets the strategic direction by engaging with European infrastructures and initiatives such as EOSC, EuroHPC, Copernicus, Destination Earth, and AI Factories. It develops a baseline ethical and legal governance framework aligned with the AI Act, GDPR, and EOSC Rules of Participation, providing compliance guidance and bias-mitigation strategies.
• WP4 consolidates this into policy recommendations, governance models, and sustainability scenarios for uptake by EOSC bodies, ERICs and policy actors.
• WP5 and WP6 design, validate, and finalise the project’s technical architecture, ensuring interoperability, FAIR-by-design principles, and seamless integration of services with EOSC.

Pillar 2 – GenAI services and scientific co-design (WP7–WP10). The second pillar focuses on developing and validating GenAI services in direct collaboration with the ESS scientific communities.

• WP7 and WP8 create AI-ready dataspaces with FAIRification, metadata enrichment, semantic search, and explainability tools. These prototypes are co-designed and iteratively refined with researchers and progressively matured into operational services.
• WP9 and WP10 validate the scientific relevance and usability of the prototypes through three real-world scientific pilots. The pilots ensure explainability, reproducibility, and EOSC alignment while driving requirements back into services and governance.

Pillar 3 – GenAI infrastructure and technical integration (WP11–WP12). The third pillar provides the federated and AI-compatible infrastructure that enables scalable and interoperable GenAI services.

• WP11 designs and prototypes the infrastructure by integrating requirements from scientific pilots and services into a federated environment combining HPC/HTC, cloud, and, if relevant, exascale resources. It also sets up essential components or core services enabling federalization such as authentication and authorisation, orchestration, and service discovery.
• WP12 deploys and matures this infrastructure, ensuring its stability, scalability, and EOSC onboarding readiness. It adopts technical system architecture from WP5 and integrates the GenAI services developed in WP7 and WP8 into production environments, supports their validation in WP9 and WP10, and establishes monitoring and sustainability mechanisms for long-term operation.

Pillar 4 – Capacity building, adoption and dissemination (WP13–WP14). The fourth pillar maximises adoption and long-term impact through capacity building and engagement.

• WP13 delivers training materials and an AI literacy toolkit for researchers and data stewards, supported by a helpdesk and quick start guides, as well as engagement campaigns, dissemination events, and open training opportunities.
• WP14 ensures effective dissemination of the resources produced in the project (FAIRification toolkits, GenAI APIs, governance frameworks, training assets) and supports end users in integrating GenAI into workflows.