Prof. Paul Yoo, Head of Data Science and AI at the University of London and director of the Birkbeck Institute for Data Science and AI (BIDA+), will deliver a seminar on May 13 at 11:00 in SPINTEC’s auditorium 445 (CEA Building 10.05, Grenoble). Attendance requires CEA entry authorization, which must be requested before May 5 via admin.spintec@cea.fr.

The talk adapts Yoo’s NATO-funded work on AI-driven defence strategy extraction to the spintronics domain. It addresses the fragility of purely data-driven forecasting models—based on graph neural networks that capture complex interdependencies among materials, device physics, manufacturing constraints, and market dynamics—when faced with technological discontinuities or policy shocks. Key to the approach are persona-engineered AI agents that engage in structured adversarial debates to challenge and recalibrate neural outputs, with epistemic scoring surfacing the most reliable strategic narratives. Explainable AI and value-chain mapping then translate these narratives into actionable levers for spintronics innovation and deployment. The resulting intelligence-to-action framework is designed to generalise beyond defence, offering robust, decision-ready strategy extraction for high-uncertainty frontier technologies.

Professor Yoo’s research spans predictive analytics, AI, and strategic forecasting. He has authored over 100 publications, secured more than US$3 million in competitive funding as principal investigator, and received awards from IEEE, Rozetta, and Samsung. His international collaborations include LSE, UCL Hospitals, KCL, and the UK Cabinet Office. He serves as editor for *ACM Computing Surveys* and *Nature (Scientific Reports)*.

On May 13th, we have the pleasure to welcome in SPINTEC Prof. Paul YOO from the University of London, United Kingdom. He will give us a seminar at 11:00 entitled : From Predictive Analytics to Actionable Strategy: Extracting Spintronics Strategies via Multi-agent Competitive Debates.

Place: IRIG/SPINTEC, auditorium 445 CEA Building 10.05 (presential access to the conference room at CEA in Grenoble requires an entry authorization. Request it before May 5th at admin.spintec@cea.fr)

Abstract: This talk extends my NATO‑funded work on extracting defence strategies via AI‑mediated competitive reasoning to the domain of spintronics, demonstrating how multi‑agent debates can transform predictive analytics into actionable technological strategy. Building on graph neural network (GNN) forecasting of complex, interconnected spintronics drivers (spanning materials, device physics, manufacturing constraints, and market dynamics) we address the brittleness of purely data‑driven models under technological discontinuities and policy shocks. Persona‑engineered agents engage in structured, adversarial debates to challenge, contextualise, and recalibrate neural outputs, with epistemic scoring used to surface the most reliable strategic narratives. Explainable AI and value‑chain mapping then translate these narratives into concrete strategic levers for spintronics innovation and deployment. The result is a generalisable intelligence‑to‑action framework that moves beyond defence applications, enabling robust, decision‑ready strategy extraction for high‑uncertainty frontier technologies such as spintronics.

Biography Paul Yoo is Professor and Head of Data Science and AI, and currently leads the Birkbeck Institute for Data Science and AI (BIDA+), University of London, United Kingdom. His research advances frontier methods in artificial intelligence, predictive analytics, and strategic forecasting, and has been recently recognised through NATO funding for pioneering big‑data‑driven forecasting methodologies. He has authored over 100 publications in leading journals and conferences and has secured more than US $3 million in competitive research funding as Principal Investigator. His work has received several prestigious awards, including the IEEE Award, the Rozetta Award, and the Samsung Award. His research portfolio features extensive international collaborations (including the LSE, UCL Hospitals, KCL, and the UK Cabinet Office), with particular emphasis on technology forecasting in areas such as defence and healthcare to support strategic decision‑making and sustainable organisational growth. He serves as an Editor for ACM Computing Surveys and Nature (Scientific Reports).

The post Seminar — From Predictive Analytics to Actionable Strategy: Extracting Spintronics Strategies via Multi-agent Competitive Debates appeared first on Spintec.