Pantelis Loupos
Assistant Professor of Marketing and Business Analytics
University of California, Davis Graduate School of Management
My research studies how interdependence changes what firms can learn and how they should make decisions. I focus on settings where customers, treatments, algorithms, information, or resources cannot be analyzed in isolation because their effects propagate through a larger system.
I combine causal inference, machine learning, network science, and optimization to study these problems in digital platforms, social networks, recommendation systems, and other technology-mediated markets. A common theme across my work is that local inference and local optimization can be misleading when decisions affect one another. My goal is to develop methods that account for these interdependencies and improve both individual-level decisions and system-wide outcomes.
Much of my applied work focuses on fintech and digital platforms, including peer-to-peer payments, crypto ecosystems, and other networked markets. I study how firms can use data and network structure to understand growth, engagement, customer value, and platform dynamics, and how they should design targeting, recommendations, and other data-driven policies.
Research Areas
Networked Digital Platforms
I study digital platforms where users are connected, behavior is visible, and value depends on social structure. My work examines peer-to-peer fintech platforms, social payment systems, digital communities, and crypto ecosystems to understand adoption, engagement, customer value, and platform risk.
Causal AI and Network Analytics
I develop and apply causal inference, machine learning, graph neural networks, and econometric methods for settings where users influence one another. A central theme in my work is how to estimate causal effects and predict behavior when outcomes are shaped by network spillovers, peer effects, and endogenous platform activity.
Platform Strategy and Decision Systems
I study how firms should use platform data to make better decisions about growth, retention, risk detection, recommendation ranking, and resource allocation. This stream connects academic research with practical problems in customer analytics, fintech, digital strategy, and AI-driven decision-making.
Selected Publications and Current Papers
- Fighting Fire with Fire: Infusing AI into Peer Review to Sustain Quality Scholarship. Joint work with Hemant Bhargava et al. Management Science, 2026.
- Social Drug Dealing: How Peer-to-Peer Fintech Platforms Have Transformed Illicit Drug Markets. Joint work with Jörn Boehnke and Ying Gu. Annals of Operations Research, 2023.
- The Shadow Value of Attention: Recommendation Ranking with Inventory Dynamics. Joint work with Can Ürgün. Under review at Management Science.
- Graph Neural Networks for Causal Inference Under Network Confounding. Joint work with Michael Leung. Revise and resubmit at The Review of Economic Studies.
Teaching
My teaching focuses on using analytical tools to improve managerial decision-making. In the MBA program, I teach core Operations Management and Customer Analytics; in the MSBA program, I teach Big Data Analytics, covering modern data systems, machine learning, network analytics, and AI. I am also a certified NVIDIA Deep Learning Institute instructor.
Contact
Pantelis Loupos
Graduate School of Management
University of California, Davis
Email: ploupos [at] ucdavis [dot] edu