This deliverable details the first release of the RI-SCALE DEP technology, and its adoption so far by the 12 validation use cases stemming from 4 RIs and 2 data spaces. Built on open standards, APIs, and protocols, the first DEP release offers baseline functionality for (1) data lifecycle management, (2) AI-based data processing and (3) access control management. The document serves as input for the next project phase, the configuration and validation of DEPs within the partner compute centers by the RI-SCALE RIs.
Executive Summary
RI-SCALE’s ambition is to enable Research Infrastructures (RIs) unlock the full value of their data holdings by coupling with computing infrastructures and offering data for AI-powered applications.
This vision is to be reached by “Data Exploitation Platforms”, scalable environments that support
- the secure transfer of data from RI data holdings to large computing resources (HPC, cloud), and
- data analysis within those systems through AI platforms, then provide model training and inference capabilities.
This deliverable details the first release of the RI-SCALE Data Exploitation Platform (DEP). This first release implements the main technical pillars for the DEP and integrates these at a proof-of-concept level, reusing technologies that have been proven in previous projects:
- Data Life Cycle Management based on Rucio, FTS, Data Preparation Service (DPS),
- AI Life Cycle management based on Model Hub and AI Agent, Interlink, itwinai, xOpat Visualisation tool, yProv4ML,
- Access Control management based on ODRL Policy Repository, Open Policy Agent (OPA), Federated tokens,
- User Credit Management based on Resource Consumption Accounting for CPU, GPU, and Storage,
- Specialised hardware integration for GROQ cards.
The primary objective of this document is to give an overview of the release, including its features and deployment guides, facilitating the next phase of the project: final configuration of the release within the 2 participating compute centres, and validation of it by the 8 scientific and 4 technical use cases. The first DEP release is not yet an integrated and ‘out-of-the-box’ software solution, therefore strong cooperation between the DEP developers, the compute centres and the 12 use cases was put in place early on during the project.
The first DEP release implements all fundamental features of the DEP, representing 23% of all the functional and non-functional requirements that have been captured from these 8+4 use cases in the first 4 months of the project1. The remaining 77% are scheduled for implementation in the second DEP release for March 2027. This deliverable details these requirements and indicates the progress towards the final DEP version.
Heimo Müller, WP5 Lead at BBMRI-ERIC
“Co-design and integrated validation are at the heart of RI-SCALE,” says Heimo Müller, WP5 Lead at BBMRI-ERIC. “By coordinating the technical work packages and running both scientific and technological validations, we ensure that the DEP reflects real-world requirements, evolves as a coherent platform rather than a collection of separate components, and delivers tangible value to Research Infrastructures.”
Ville Tenhunen, RI-SCALE Technical Coordinator and Data Solutions Architect at EGI Foundation
“Importantly, this first release is not yet a fully integrated, out-of-the-box product,” says the RI-SCALE Technical Coordinator Ville Tenhunen. “It represents a structured and validated baseline. We have built it in close co-design with the use cases and compute centres to ensure that every technical step is grounded in real research needs. The lessons we learn during validation will directly shape the second DEP release in March 2027, where we will deliver the remaining requirements and significantly mature the platform’s integration, user experience, and sensitive data handling capabilities.”
