On March 4, 2025, EuroCC Denmark, in collaboration with AI-Boost and DWARF, hosted a highly anticipated workshop titled "Scaling AI Solutions: From Workstations to Supercomputers" at DWARF in Frederiksberg. The event aimed to provide participants with practical tools to effectively use MLOps and ensure their AI projects support business goals, future-proofing AI investments and maintaining a competitive advantage in a rapidly evolving world.
The workshop kicked off with a warm welcome from Mads Bie, Business Innovation Director at Dwarf, followed by an introduction to the day's program by Dennis Wollbrink from EUROCC2 and Signe Ingholt-Gaarde from AI-Boost. The agenda was packed with insightful sessions led by industry experts, each focusing on different aspects of MLOps, AI and supercomputing.
The workshop showcased cutting-edge advancements in AI through machine learning processes, emphasizing key themes that resonate across diverse industries and applications. One of the central ideas was the importance of streamlining the AI development lifecycle through robust MLOps practices. As Nicki Skafte Detlevsen, Post doc at DTU Compute pointed out, MLOps bridges the gap between machine learning, IT operations, and DevOps, ensuring that AI systems are developed efficiently and maintained with minimal technical debt. This approach fosters interdisciplinary collaboration and safeguards the longevity of AI projects.
Another recurring insight was the necessity of practicality and domain expertise in applying AI solutions to real-world problems. Kasper Honoré, Technical Project Manager at FORCE Technology, illustrated this with compelling examples, such as predictive maintenance in industrial settings and innovative monitoring techniques. Whether optimizing refinery equipment or preventing failures in grass-cutting machines, simplicity and cost-effectiveness emerged as critical factors in successful implementations.
Supercomputing capabilities add a new dimension to scaling AI solutions, as highlighted by Julius Roeder's, HPC Consultant at DEIC, exploration of tools like the LUMI supercomputer. By leveraging the unparalleled speed and power of supercomputers, organizations can run complex experiments, scale models, and expedite workflows faster or more efficient. These benefits were contrasted with cloud-based approaches, offering participants a clear understanding of when and how to adopt supercomputing technologies effectively.
Finally, the transformative potential of machine learning in specialized fields, such as weather prediction, demonstrated its far-reaching impact. Simon Kamuk Christiansen, Machine Learning Researcher at DMI (National Danish Meteorological Institute). showcased how machine learning could drastically reduce computation time while maintaining physical consistency, enabling real-time predictions and higher resolution models. This innovation paves the way for enhanced decision-making in critical areas like climate forecasting.
In a final session, Niklas Hall, business developer at the Copenhagen Business Hub, emphasized the importance of starting AI projects with clearly defined business problems and minimizing risks through knowledge and collaboration. He highlighted the AI Boost program as an important resource for identifying challenges, testing models with expert support and using supercomputing tools to effectively advance AI initiatives.
The workshop concluded with a wrap-up session led by Dennis Wollbrink from EUROCC2 and Signe Ingholt-Gaarde from AI-Boost, who discussed the possibilities for next steps. Participants were encouraged to network, reflecting on the day's insights and exploring opportunities for collaboration.
Overall, the workshop provided concrete insights into how MLOps can drive innovation, minimize risk, and optimize the use of resources. Attendees left with a clear actionable plan customized to their business needs, ready to take their AI projects to the next level with advanced computing power.
For more information and to stay updated on future events, visit the EuroCC Denmark website and follow @EuroCC Denmark at LinkedIn.
