EMPYREAN participated in Data Week 2026, held on 5–6 May 2026 in Oslo, Norway, contributing to the workshop “From Data to Decisions: AI Lifecycle Management Across the Cognitive Computing Continuum.” The workshop brought together European research and innovation projects working on distributed AI, cognitive computing, and the Edge–Cloud continuum.
The session, organised within the European Cognitive Computing Continuum community, focused on the challenges of managing the complete AI lifecycle across distributed and heterogeneous computing infrastructures. Discussions covered topics including AI pipeline orchestration, federated learning, MLOps and AIOps, distributed data management, monitoring, cybersecurity, trustworthiness, privacy and energy-efficient AI services.
Representing EMPYREAN, Panagiotis Kokkinos (ICCS/NTUA) presented the project’s approach to adaptive orchestration frameworks, SDKs and distributed MLOps toolchains for operationalising AI services across heterogeneous edge-to-cloud environments. The presentation highlighted the challenges of dynamically scheduling and optimising AI workloads across infrastructures with different compute, latency and energy characteristics, while maintaining efficient operation and observability.
The workshop brought together contributions from several Horizon Europe projects, including EMPYREAN, CoGNETs, ENACT, INTEND and Swarmchestrate, providing an opportunity to exchange knowledge and identify synergies around the development of the Cognitive Computing Continuum. The session also highlighted the complementary approaches being developed across the projects, ranging from decentralised coordination and swarm intelligence to adaptive orchestration, intent-driven infrastructure management and energy-aware optimisation.
For EMPYREAN, the workshop provided an opportunity to further discuss how autonomous and adaptive orchestration can support AI applications across the Edge-Cloud continuum, and how heterogeneous resources can be coordinated to meet the dynamic requirements of next-generation applications and services.
EMPYREAN thanks the organisers and participating projects for the valuable discussions and the opportunity to contribute to this exchange on the future of AI lifecycle management and cognitive computing infrastructures.
