RI-SCALE has reached a pivotal point in its mission to co-design, prototype, and validate Data Exploitation Platform (DEP) technology by achieving the first  three technical milestones this spring.

Data Orchestration Live 

The first major achievement, Milestone 2.1, was reached at the end of March with the deployment of the DEP’s Data Orchestration Layer. By integrating EGI Check-in service with Rucio and the File Transfer Service (FTS), the project has established a secure and federated environment for scientific collaborators.

With the initial Research Infrastructure users now onboarded, the service is available to begin the integration and validation of scientific data required for upcoming AI-driven pilot cases.

“We at CERN are very happy with this new development in the context of the RI-SCALE project, because it demonstrates how the powerful data management tools we’ve built for high-energy physics can be successfully adapted for a much broader scientific community,” says Martin Barisits from CERN. “By providing this orchestration layer, we ensure that researchers across Europe can manage massive datasets without needing to be IT experts themselves.”

Scaling AI: HPC Interoperability

Milestone 3.1 was reached in April, establishing interoperability between core software tools including iTwinAI, interLink, yProv4ML, and AI Model Hub. A major highlight was the deployment of a JupyterHub instance, enabling researchers to offload heavy computational jobs from the cloud directly to HPC nodes. This capability was first validated by the ENES climate research infrastructure use case and is now being scaled to other RI-SCALE research infrastructures  like EISCAT, EuroBioImaging, and BBMRI.

Secure Access and Accounting Framework

Finally, Milestone M4.1 was achieved with the deployment of the Access Management Frameworks across the RI-SCALE DEP environments. This policy-based authorisation framework, utilising the ODRL Policy Repository, OPA-based Policy Decision Point, and OpenID Federation Trust Anchor, establishes secure trust relationships across participating services. Initial integration with the Credit Management System (CRMS/ARGO Accounting) is already underway, enabling the platform to authenticate and record resource usage. The framework is now available for testing and validation by RIs in line with the planned DEP release.

Why Does It Matter?

For most researchers, the logistical difficulty of moving data to powerful computing sites is often a significant barrier in their work. This development in the project means that data movement and high-performance computing are now unified into a single environment. By removing the friction of fragmented infrastructure, RI-SCALE ensures that AI-driven research data analysis can finally be executed at scale.

The Next Steps

With the development of data orchestration, AI interoperability, and secure access now established, the RI-SCALE project enters its next phase of validation.

“Next, the project will focus on validation through 12 scientific use cases and 4 technical use cases. Meanwhile, technical development will continue, with increasing emphasis on the second release of the DEP, which will incorporate most of the functionality, as well as internal and external interoperability with research infrastructures (RIs), other AI-related infrastructures, and more. This also involves deepening relationships with the validation use cases,” says RI-SCALE technical coordinator Ville Tenhunen.

The project will also address policy development, training needs, industry engagement, and ethical assessments of the DEP workflows and operating environments. Additionally, the project is seeking new partners to co-design, validate, and benefit from the Data Exploitation Platforms (DEPs) being developed.