Research

I study how firms can design learning systems that optimize not only predictive accuracy, but also the quality of the downstream decisions they support. This line of research combines machine learning, data-driven optimization, and statistical modeling to align learning with firms' economic objectives.

• Decision-Aware Segmentation (Working Paper)
  Joint work with Sandeep Chandukala (SMU), Ernst Osinga (SMU), and Spyros Zoumpoulis (INSEAD)
  
  This paper applies a decision-aware learning perspective to segmentation for targeting, aligning segmentation with the firm's downstream economic objectives. Standard segmentation methods optimize statistical cluster fit, not downstream targeting performance. Instead of treating customer segmentation as a descriptive tool, we ask: how should firms learn decision-relevant customer segments that improve targeting outcomes?

I study how human and AI can jointly advance scientific knowledge production.

• A Human-AI Collaborative Framework for Theory Building (Working Paper)
  Joint work with Philip Parker (INSEAD), Phanish Puranam (INSEAD), Eric Luis Uhlmann (INSEAD), and Spyros Zoumpoulis (INSEAD)
  
  We ask how human scholars and AI can jointly build theory. Using discourses on race and gender inequality in a large corpus of academic papers, we compare their forecasts and explanations against machine-learning evidence—and show that each intelligence contributes distinct, complementary strengths to credible scientific knowledge production.