Online learning
Algorithms with finite-time guarantees for sequential decisions in uncertain and strategically changing environments.
Doctoral researcher · Georgia Tech
Learning to decideunder uncertainty.
I study the mathematical foundations of online learning and decision-making—especially when the environment is uncertain, adaptive, or adversarial.
I am a Ph.D. student in Electrical and Computer Engineering at the Georgia Institute of Technology, advised by Prof. Matthew Hale and co-advised by Prof. Dmitrii Ostrovskii. My work connects online optimization, optimal experimental design, and multi-agent systems.
Based inAtlanta, Georgia
AffiliationCORE Lab · Georgia Tech
Research focus
Principled learning for interactive systems.
I build algorithms and theory for decisions that unfold over time, where information is limited and other agents may respond.
Experimental design
Adaptive methods for choosing informative experiments while balancing geometry, computation, and regret.
Multi-agent systems
Learning in games where agents adapt, objectives evolve, and interaction itself shapes the data we observe.
Selected work
Publications & current research.
My recent work develops learning methods for interactive, sequential environments, building bridges between regret theory, experiment design, and strategic decision-making.
Conference paper
Online Scalarization in Vector-Valued Games
IEEE Conference on Decision and Control (CDC), 2026
A bi-level online learning framework that treats scalarization as a decision variable, using adaptive objectives to shape outcomes in repeated vector-valued games.
Working paper
Online D-Optimal Experimental Design
In preparation
Online algorithms for log-determinant design losses, with an emphasis on efficient variational updates and regret guarantees that reveal the geometry of experimental design.
In progress
Journal paper
Optimal Management for Megawatt Level Electric Vehicle Charging Stations With a Grid Interface Based on Modular Multilevel Converter
IEEE Access, vol. 10, 2022
An optimization-based load-management strategy for large-scale charging facilities, validated through daily-operation and real-time simulations.
News & highlights
Recent milestones.
Online Scalarization in Vector-Valued Games was accepted to IEEE CDC 2026.
Released the preprint Online Scalarization in Vector-Valued Games on arXiv.
Presented Risk Domination via Adaptive Scalarization in Vector-Valued Games at the Georgia Tech DCL Student Symposium.
Teaching & service
Research is a conversation.
International TA Liaison
I support international teaching assistants at Georgia Tech as they navigate teaching, communication, and academic life across cultures.
Georgia Tech profileOpen to collaboration
I enjoy discussing online learning, optimal design, adaptive adversaries, and multi-agent decision systems with researchers across control, optimization, and machine learning.
Start a conversation