I am a fourth-year PhD candidate in Operations Research at Georgia Tech working with Professor Alejandro Toriello. My research focuses on developing practical algorithms with theoretical guarantees for dynamic matching and scalable algorithms for large-scale logistics optimization. I am also broadly interested in optimizing dynamic and stochastic systems. In summer of 2024, I interned in middle mile planning and routing optimization science team at Amazon and worked on driver break modeling in large-scale middle-mile routing problem. Prior to joining Georgia Tech, I received my BS and MS in Industrial and Management Engineering from Pohang University of Science and Technology (POSTECH), where I worked on scheduling and transportation problems in manufacturing and healthcare.
CV (last updated March 2025)
Upcoming talks
- 2025 INFORMS Computing Society Conference
- Presentation title: Batching and Greedy Policies: How Good Are They in Dynamic Matching?
- Session: Stochastic and Dynamic Optimization for Emerging Supply Chain and Transportation Systems
- Date: Friday, March 14 | 08:30 AM – 10:00 AM
- Location: Music Room at the University of Toronto
- IISE Annual Conference & Expo 2025
- Presentation title: Recursive Partitioning and Batching for Network Design with Service Time Guarantees at Massive Scale
- Date and Location: TBD
Research Projects
- Asymptotic optimality of batching and greedy policies in dynamic matching (with Dr. Alejandro Toriello)
- We study a dynamic non-bipartite stochastic matching problem, motivated by ride-sharing and freight transportation marketplaces. We analyze the asymptotic optimality of batching and greedy policies and show how fast they converge to the optimal. Interestingly, both policies can achieve near-optimal performance with exponential convergence, even when considering impatient nodes.
- This work highlights how simple, well-structured policies can efficiently balance system performance and maximum waiting time.
- Dynamic matching under node impatience (with Dr. Alejandro Toriello and Dr. Diego Cifuentes)
- Our previous study shows that simple batching and greedy policies may not be optimal when nodes are highly impatient. We aim to develop a stronger relaxation model for a dynamic non-bipartite stochastic matching problem under node impatience.
- Recursive partitioning and batching for network design with service time guarantees at massive scale (with Dr. Alejandro Toriello and Dr. Alan Erera)
- We study a large-scale network design problem, motivated by parcel delivery industry. To handle the complexity of real-world instances, we propose a recursive graph partitioning and batching method, which breaks the problem into smaller regions and effectively merges the solutions for each subproblem.
- Our method finds high-quality solutions within hours for a real-world instance with over one million arcs and 40,000 commodities, while a commercial solver is unable to even build a model for a much smaller instance.