My name is Zixuan, and I am a Research Associate at the University of Bristol’s AI for Collective Intelligence (AI4CI) Hub. My research sits at the intersection of artificial intelligence, collective behaviour, and public health, and I am particularly interested in how we can design models that help people make better decisions under uncertainty.
I completed my PhD in Engineering Mathematics at Bristol, where I investigated how groups of agents learn and adapt when facing incomplete information. During this time, I became interested in how uncertainty shapes collective behaviour and how mathematical modelling can provide insight into complex, dynamic systems. That experience sparked a broader interest in problems that combine rigorous mathematical modelling with real-world complexity. Over time, I have become especially motivated by questions that matter not just to theory, but also to society – how can AI and data science be used responsibly to support decision-making in high-stakes contexts like public health?
What are you working on?
My current work focuses on developing AI tools that can strengthen pandemic preparedness and resilience. At the core is the question of how we can design systems that remain useful when information is uncertain, incomplete, or rapidly evolving.
More specifically, I am working on a prototype “AI agent” that integrates different forms of evidence—quantitative data, expert knowledge, and stakeholder perspectives—and translates them into transparent, interpretable recommendations. This approach goes beyond forecasting disease trajectories: it aims to help policymakers explore intervention strategies, compare trade-offs, and engage in evidence-based dialogue.
I am always keen to connect with researchers, practitioners, and public health partners interested in testing, refining, and applying these tools in real-world contexts.
What excites you most about your field of research?
What excites me most is the opportunity to use AI to help people make better decisions when the stakes are high and the answers are uncertain. During my PhD, I became fascinated by how groups learn and adapt, and I now see those same dynamics play out in public health. I enjoy building systems that don’t just predict outcomes, but also make evidence clearer and more usable for the people who need it most. Knowing that my work could one day influence real choices and safeguard lives is what motivates me every day.
Who has influenced your academic journey?
My academic journey has been guided by mentors and collaborators who have shaped both my thinking and approach to research. During my PhD, my supervisors, Jonathan Lawry and Michael Crosscombe encouraged me to pursue ambitious ideas and helped me develop the detailed mathematical and computational skills to realise them.
I am also grateful for collaborations with Oliver Y Chen at Lausanne University Hospital and the University of Lausanne, and Bangdong Zhi at the University of Bristol. Working with them inspired me to apply my background more directly to healthcare challenges. More recently, my line manager, Leon Danon, has inspired me to think beyond methods and focus on how modelling can directly inform public health decisions.
Can you recommend a paper you think we should read?
I recommend Epidemic Modeling with Generative Agents by Williams, Hosseinichimeh, Majumdar, and Ghaffarzadegan (2023) . The paper introduces a new paradigm for epidemic modelling where each agent is powered by a generative AI, allowing them to reason and make decisions based on evolving information. Remarkably, these agents exhibit behaviours such as self-isolation and risk-responsive mobility, leading to epidemic waves and curve-flattening effects. I find it exciting because it opens new possibilities for integrating human behaviour, AI, and epidemiology in dynamic and interpretable ways.
What are your ‘Top Tips’ for early career researchers?
One of the most important lessons I’ve learned is to treat research as a privilege rather than a chore—enjoy the process, because it’s one of the few times in our career where the work is entirely your own. We’ll encounter dips, but these are part of the journey: when it happens, change the angle, try a new method, or draw inspiration from a fresh paper. Always validate and verify the work carefully before moving on; small mistakes can unravel later. And don’t forget the human side: build networks, share knowledge generously, and develop resilience when facing criticism. Successes may feel rare day-to-day, but over time they accumulate—and they are worth celebrating.