The deliverable outlines the architecture of the EOSC Data Commons and its services
EOSC Data Commons Delivers Its First Production-Ready Platform Release

D3.2 “EOSC Data Commons – First Software Release” marks the transition from architectural design and prototypes to a fully integrated, production-ready EOSC Data Commons platform.
This first software release provides a working environment where researchers can discover, access, combine, analyse, and reuse data across heterogeneous repositories and computing infrastructures.
From Design to Operational Platform
Building on the EOSC Data Commons framework and earlier Proof of Concept work around the Dispatcher and EOSC Data Player, this release moves the project from concept to an operational service ecosystem for real research workflows. The deliverable documents how the core services of EOSC Matchmaker and EOSC Data Player have been deployed together to support end-to-end research journeys, from dataset discovery to tool execution in distributed Virtual Research Environments (VREs).
What's in Release 1
Release 1 provides the first production deployment of the EOSC Data Commons architecture, integrating software components and databases into a common operational environment.
The platform now supports metadata aggregation for 10 repositories across 26 harvesting endpoints, enabling researchers to search, match and orchestrate data and tools at scale.
Key capabilities introduced with this first software release include:
- Integrated EOSC Matchmaker and EOSC Data Player services, supporting seamless progression from data discovery to execution of analytical tools and workflows in distributed VREs.
- A Metadata Warehouse with an extract–transform–load pipeline that harvests, normalises, enriches and indexes metadata from diverse sources, powering advanced search and matchmaking.
- OpenSearch- and large language model-based dataset discovery, combining semantic embeddings and AI-assisted search to help users find relevant datasets and tools using natural language queries.
- File-level metadata management (fileDB) and tool metadata via the Tool Registry and Tool Metadata Harvester, enabling precise matching between datasets, files and computational tools.
- A Coordinator and RO-Crate Creator that packages data, tools and workflows into research objects, ready to be dispatched to target VREs for execution.
Architecture, Infrastructure and Operations
D3.2 documents the deployment of 15 internal software components and five supporting databases that together implement the EOSC Data Commons Release 1 architecture.
These components cover metadata crawling, transformation, scheduling, search, user interface, file and tool metadata management, coordination, and dispatch to execution environments.
The production platform is deployed using containerised services and Infrastructure-as-Code practices, ensuring reproducible environments and streamlined operations.
Monitoring and observability are provided through Prometheus, Grafana, Loki, Alertmanager and complementary dashboards, while user authentication and authorisation are handled via EGI Check-in.
A Foundation For Future EOSC Service Expansion
With the first production release in place, EOSC Data Commons now has a stable foundation to expand services, integrate additional repositories and onboard more VREs across disciplines.
Future development will focus on improving metadata quality and search relevance, enhancing AI-assisted discovery, broadening tool interoperability and increasing automation of research workflows to support a growing EOSC user community.

