17 June 2025 · 1 min read
Some good learnings from the AIspotlight at HLTHEurope2025 on federated data and learning coll...
I share key learnings from the AIspotlight at HLTHEurope2025 on federated data and learning collaboration, where speakers from Roche, Charité, and beyond made the case for combining insights across institutions. A recurring theme: this shift in how we interact with health data requires clearer explanation and specific use cases for stakeholders.
Some good learnings from the AIspotlight at HLTHEurope2025 on federated data and learning collaboration, with David Champeaux, Ariel Dora Stern, Dr. Florentine Kaniess (b. Radelfahr), and Frederik Buijs.
Roche has been using federated data models for quite a while now as individual academic sites just have seen that you need to combine the insights from multiple sources.
Even a large academic center like Charité - Universitätsmedizin Berlin admits the need for combining data into larger federated sets. As one example, Florentine quoted a collaboration of most large German university hospitals on big data, as only combined these institutions can compete with the large Ivy League institutions in the US.
From the perspective of Ariel Dora Stern this represents a mindset shift in how to interact with health data. But this mindset shift needs more explanation to stakeholders, and be specific about the use cases.
Related insights
4 Nov 2025
Pharma used to rent compute. Eli Lilly and Company just bought the factory.
7 Apr 2026
Anthropic buys Coefficient Bio: why the scientific decision layer matters
19 Mar 2026
Microsoft Copilot Health: why launch architecture is not execution
10 Mar 2026
AI drug discovery benchmarks: why pharma needs model evaluation discipline