I build and evaluate AI systems - behavioural fine-tuning, multi-agent architectures, and agent-based simulation at scale. My work sits at the intersection of AI and the social sciences, with applications in human behaviour, policy, and international law.

Currently

  • Postdoctoral Research Assistant - Royal Holloway, University of London (Leverhulme Trust)
  • Research Associate - King’s College London (part-time)
  • Tutor - Complexity72h, London, June 2026

Recent Work

Luckner, K., Prabhakaran, A., Kos, U., Küçüksu, A., Lundsgaard, T., Fikfak, V. (2025). Text, Interviews, and Expert Coding: Using Qualitative and Quantitative Data in an Agent-based Model of Compliance with Human Rights Judgments. Advances in Social Simulation. Springer.

Luckner, K., Prabhakaran, A., Fikfak, V., Ridi, N. (2025). Why the Study of International Law Needs Agent-based Modelling. Under review, Journal of International Dispute Settlement.

Duran-Sala, M., et al., Prabhakaran, A. (2024). Disentangling individual-level from location-based income uncovers socioeconomic preferential mobility and impacts segregation estimates. Under review, Nature Communications.

All publications →