Quality Assurance in HPC/AI Training

Institutions engaged in High Performance Computing (HPC) and AI (Artificial Intelligent) face the same underlying challenge: how do we ensure quality, consistency and learner engagement in rapidly expanding training ecosystems? Institutions across Europe and globally are confronted with similar issues worldwide and so this Birds-of-a-Feather (BoF) session brings these communities together to articulate their questions, compare approaches and collectively deepen their judgment on these topics.

We represent a broad group of organisations delivering and coordinating training across Europe: national competence centres, supercomputing centres, universities and EuroHPC-funded projects such as EVITA, AI Factory, CASTIEL2 and EuroCC2. We operate within similar constraints and expectations: designing and delivering courses for diverse and international audiences, maintaining quality over time, and aligning practices across teams, countries and project frameworks. As practitioners responsible for training tasks in large European collaborations, we all confront the same need to harmonise processes, establish shared standards and ensure that QA supports, not restricts, the development of impactful and scalable training ecosystems.

The session will introduce the key subjects of Quality Assurance (QA) for discussion:

- One key issue is consistency in training design and delivery. As the number of courses grows, shared guidelines, templates and minimum structural requirements become essential to providing a coherent learner experience. A stable didactic framework, learning outcomes, prerequisites, module organisation, assessment expectations, creates a reliable foundation while still allowing trainers the flexibility to adapt materials to their context.

- A second theme is the verification of content quality and pedagogical soundness. In fast-evolving domains such as HPC and AI, the accuracy, relevance and methodological appropriateness of training materials need to be continuously reassessed. QA involves systematic checks to ensure that content is current, technically correct and aligned with the target audience. Institutions increasingly rely on internal reviews, expert validation and structured update cycles to maintain credibility and trust.

- A third key dimension concerns evaluating training effectiveness and enabling continuous improvement. Surveys, short assessments and follow-up evaluations help measure satisfaction, learning progress and the real-world application of skills. Trainer reflection and peer exchange are equally essential to further sustain training quality improvement. QA is therefore less a static checklist than a dynamic process embedded in the training lifecycle.

- Finally, the BoF session will address practical questions that many trainers face in their daily work. How do we ensure engagement in virtual classes where body language and visual cues are absent? How can we adapt interventions to sustain attention across cultural or disciplinary contexts? These issues, though subtle, are central to QA because they directly shape learner experience and satisfaction.

This BoF invites all participants to share their perspectives, raise unresolved questions, and reflect together on the strategic and practical dimensions of QA in HPC and AI training. The goal is not to impose a single framework, but to identify common principles and support a shared understanding of what high-quality, engaging and trustworthy training should look like.

---

Organizers:

• Sima Barzegar
• Maria-Ribera Sancho