Sean Lo

About

Hello! My name is Sean, and I am a third-year PhD student in Operations Research at Massachusetts Institute of Technology. I am very fortunate to be advised by Prof. Alexandre Jacquillat. I am interested in large-scale integer optimization and sequential decision-making under uncertainty, with applications in transportation, healthcare, and logistics.

In my free time, I enjoy running, playing the horn, and climbing. I grew up in Singapore.

Email: seanlo [at] mit [dot] edu
Google Scholar | GitHub | LinkedIn

News

(Jun 2025) New preprint on online rack placement in data centers at Microsoft! This features an online resource allocation / bin packing structure over a long time horizon. We use a sampling-based resolving heuristic and demonstrate real-world implementation and benefit.
(May 2025) Started a research internship at Microsoft Research with the Machine Learning and Optimization group! 📍Redmond, WA
(May 2025) Updated preprint for low-rank matrix completion! New experiments demonstrate greater scalability of our exact method for a variety of instances.
(Mar 2025) Gave a talk at ICS in Toronto on an online assignment + routing problem in organ delivery! This is an ongoing collaboration with a healthcare startup in MA.

Research

(Author order is alphabetical.)

Online Rack Placement in Large-Scale Data Centers: Online Sampling Optimization and Deployment

Saumil Baxi, Kayla Cummings, Alexandre Jacquillat, Sean Lo, Rob McDonald, Konstantina Mellou,
Ishai Menache, Marco Molinaro
Jun 2025 (preprint)

We investigate the problem of placing incoming racks of servers in a datacenter, given space, power, and cooling restrictions. This problem features a online resource allocation / bin-packing structure, for which we propose a tractable online sampling optimization (OSO) approach to multi-stage stochastic programming -- approximating unknown parameters with a single sample path and re-optimizing dynamically. We show theoretical guarantees on OSO's performance on canonical online resource allocation and bin packing problems, and demonstrate the impact of deployment across Microsoft's fleet of datacenters.

Subpath-Based Column Generation for the Electric Routing-Scheduling Problem

Alexandre Jacquillat, Sean Lo
Jul 2024 (preprint) (code)🏅 Finalist: INFORMS TSL Best Student Paper Award 2024

We investigate the classic VRP, with a twist that vehicles are electrified and have to stop at charging stations to recharge. This problem has various applications: last-mile logistics with electric vans, drayage via electric trucks, micromobility with electrified light vehicles, and drone delivery in urban settings. While this problem has been extensively studied in the literature, our approach is unique because we study the more challenging variant in continuous time and charge. We have a path-based formulation for the routing problem (as usual), and we develop a column generation algorithm customized for the electrified nature of our problem. Our algorithm decomposes the dynamic program into two stages: generating subpaths between charging stations, and stitching subpaths together to form complete paths from depot to depot. Our new subpath-based domination framework cleanly accommodates state-of-the-art improvements in vehicle routing, such as ng-route relaxations and limited memory subset-row cuts. Computational experiments show that our method clearly outperforms the current best method and leads to significant cost savings over optimizing routing and charging decisions separately.(This work was recognized as one of four finalists of the INFORMS TSL Best Student Paper Award in 2024!)

Disjunctive Branch-And-Bound for Certifiably Optimal Low-Rank Matrix Completion

Dimitris Bertsimas, Ryan Cory-Wright, Sean Lo, Jean Pauphilet
May 2023 (this version May 2025) (preprint) (code) (experiments)

We consider the problem of finding a low-rank matrix that is optimally fits observed entries, minimizing the sum of MSEs (applications include compressed sensing and recommender systems). We build a branch-and-bound algorithm for this problem using various ingredients: a semidefinite relaxation of the low-rank constraint, a branching scheme that uses eigenvector disjunctions, and an alternating minimization method that yields high-quality solutions. We show that our method outperforms existing exact methods, and show scalability to larger instances (including rectangular matrices).(I worked on this during my Masters' program!)

Presentations

Online Assignment-Routing for Organ Delivery

Talk at INFORMS Computing Society Conference 2025
Mar 2025, Toronto, Canada

Subpath-Based Column Generation for the Electric Routing-Scheduling Problem

Talk at INFORMS Annual Meeting 2024
Oct 2024, Seattle WA, USA
Talk at International Symposium for Mathematical Programming 2024
Jul 2024, Montreal, Canada
Talk at INFORMS Optimization Society Conference 2024
Mar 2024, Houston TX, USA
Talk at INFORMS Annual Meeting 2023
Oct 2023, Phoenix AZ, USA

Disjunctive Branch-And-Bound for Certifiably Optimal Low-Rank Matrix Completion

Poster at Mixed-Integer Workshop 2023
May 2023, Los Angeles CA, USA

Teaching

MIT 15.C57 Optimization Methods

Teaching Assistant, Fall 2024Teaching assistant for graduate class with over 250 students. Developed material, conducted recitations, graded assignments, projects and exam. (evaluations)

MIT 15.S60 Optimization with Julia and JuMP

Session Instructor, Jan 2024Session instructor for a practical optimization session for first-year doctoral students.

MIT 15.083 Integer Programming and Combinatorial Optimization

Teaching Assistant, Spring 2023Sole teaching assistant for doctoral class with over 30 students. Developed material, conducted recitations, graded assignments, projects and exam. (evaluations)

Education

Ph.D. in Operations Research

Massachusetts Institute of Technology, Operations Research Center, 2022 - 2027 (expected)
Advisor: Alexandre Jacquillat

Master in Business Analytics

Massachusetts Institute of Technology, Sloan School of Management, 2021 - 2022

Bachelors of Science in Mathematics

Imperial College London, 2017 - 2020

Professional Experience

Research Intern

Microsoft Research: Machine Learning and Optimization Group
Summer 2025

Service

Student Coordinator: MIT ORC Seminar Series, Fall 2024

Invited academics to speak at MIT ORC's seminar series, and made logistical arrangements.

Page updated

Google Sites

Report abuse