About

Divya Singhvi is an Assistant Professor of Technology, Operations, and Statistics at the NYU Stern School of Business, having joined the institution in September 2021. His research focuses on the intersection of machine learning and operations management, specifically addressing problems related to optimal demand learning, pricing, recommendations, and logistics within online and offline retail operations. Prior to his current role, Singhvi spent a year at IBM Research as a postdoctoral researcher in the IBM Research AI Residency program. He holds a Bachelor of Science degree in Operations Research and Engineering from Cornell University and a PhD in Operations Research from the Massachusetts Institute of Technology.

Research topics

• Computer science
• Business
• Operations research
• Machine learning
• Econometrics

Selected publications

• Task Assignments in Distributed Supply Chains When Downtime Hurts: A Data-Driven Approach

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access
  PublisherDOI

• Online Learning with Sample Selection Bias

  Operations Research · 2025-03-19

  article1st authorCorresponding

  Personalized recommendation systems often face the challenge of making optimal decisions when user preferences are unknown, and outcomes are only observed if users engage with the platform (e.g., clicking a recommendation). In their paper, “Online Learning with Sample Selection Bias,” Singhvi and Singhvi study this problem in the context of sequential decision making, where the censoring of outcomes leads to selection bias. Ignoring this bias results in suboptimal recommendations and linear regret, even for well-performing existing learning algorithms. To address this, they propose the sample selection bandit (SSB) algorithm, which combines Heckman’s two-step estimator with the “optimism under uncertainty” principle. The authors also demonstrate that SSB achieves a near-optimal regret rate. Extensive numerical experiments using synthetic and real-world donation data confirm that SSB significantly outperforms existing algorithms, effectively addressing selection bias while improving recommendations and outcomes in practical settings.

  PublisherDOI

• Deep Policy Iteration with Integer Programming for Inventory Management

  Manufacturing & Service Operations Management · 2025-01-06 · 10 citations

  articleSenior author

  Problem definition: In this paper, we present a reinforcement learning (RL)-based framework for optimizing long-term discounted reward problems with large combinatorial action space and state dependent constraints. These characteristics are common to many operations management problems, for example, network inventory replenishment, where managers have to deal with uncertain demand, lost sales, and capacity constraints that results in more complex feasible action spaces. Our proposed programmable actor RL (PARL) uses a deep-policy iteration method that leverages neural networks to approximate the value function and combines it with mathematical programming and sample average approximation to solve the per-step-action optimally while accounting for combinatorial action spaces and state-dependent constraint sets. Methodology/results: We then show how the proposed methodology can be applied to complex inventory replenishment problems where analytical solutions are intractable. We also benchmark the proposed algorithm against state-of-the-art RL algorithms and commonly used replenishment heuristics and find that the proposed algorithm considerably outperforms existing methods by as much as 14.7% on average in various complex supply chain settings. Managerial implications: We find that this improvement in performance of PARL over benchmark algorithms can be directly attributed to better inventory cost management, especially in inventory constrained settings. Furthermore, in the simpler setting where optimal replenishment policy is tractable or known near optimal heuristics exist, we find that the RL-based policies can learn near optimal policies. Finally, to make RL algorithms more accessible for inventory management researchers, we also discuss the development of a modular Python library that can be used to test the performance of RL algorithms with various supply chain structures. This library can spur future research in developing practical and near-optimal algorithms for inventory management problems. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.0617 .

  PublisherDOI

• Buying Cheap: Brand Switching During Economic Distress and Its Disparate Impact on Consumers

  Manufacturing & Service Operations Management · 2025-02-26 · 3 citations

  article

  Problem definition: Improving the surplus of low-income consumers during economic distress is of primary concern for many governments. This paper uses an economic model to investigate consumers’ brand switching during economic distress and highlights its disparate impact on low-income consumers. Our modeling framework also captures how retailers and national-brand manufacturers strategically adjust the market prices in response to brand switching and shelf space constraints. To generate prescriptive insights, we also analyze the effectiveness of commonly observed government interventions that fall into two major categories: (i) consumer-focused (e.g., cash subsidy) and (ii) retailer-focused (e.g., price control). Methodology/results: Our analysis indicates that market access for low-income consumers can decline significantly because of the brand switching behavior. Furthermore, not all government interventions are equally effective in increasing the welfare of low-income consumers. In fact, retailer-focused schemes such as price control (PC) can backfire and decrease access for low-income consumers. Although cash subsidy (CS) can increase the surplus of low-income consumers, it is always at the expense of high-income consumers. Managerial implications: We study an important but understudied challenge that highlights how strategic behavior by retailers can exacerbate affordability and accessibility concerns during economic distress. Model calibration using NielsenIQ Homescan Panel data shows that our model captures the data well and generates practical insights for policymakers. These results suggest that the success of government interventions depends critically on whether they account for the strategic behavior of different stakeholders in the supply chain. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.0380 .

  PublisherDOI

• Design of Resale Platforms

  2025-07-02

  articleOpen access

  We study resale platforms, an emerging type of online marketplaces in developing countries. Resale platforms are designed for individuals (resellers) to sell products to others as opposed to buying for themselves, enabling them to supplement their income by earning a margin on the transactions they generate. One challenge these platforms face is that competition among resellers may emerge as more of them join the platform, as their social circles increasingly overlap.

  PublisherOA PDFDOI

• A Data-Driven Approach to Improve Artisans’ Productivity in Distributed Supply Chains

  Operations Research · 2025-10-27 · 1 citations

  article1st authorCorresponding

  Smarter Supervision Lifts Rural Weavers’ Productivity Frequent, predictable supervisor visits can be a powerful lever for boosting artisan productivity in distributed supply chains, according to a study conducted with Jaipur Rugs in India. Analyzing loom-level data, the authors show that reducing the average gap between visits by just one day raises weaving rates by 8.5%—with more substantial gains on complex rugs and when visits follow consistent schedules. Building on these insights, they develop a routing and scheduling framework that targets those looms most in need of support. In a 25-week field implementation covering about 6,000 visits across 200 looms, sites assigned to optimized routes saw a 16.7% increase in weaving speed relative to controls, highlighting a practical, low-cost pathway to higher earnings for rural women weavers. The research suggests that data-driven supervision in other supply chains with a similar structure (e.g., smallholder agriculture) could boost productivity and earnings, offering an operational lever for poverty alleviation at scale.

  PublisherDOI

• A Data-driven Approach to Improve Artisans’ Productivity in Distributed Supply Chains

  2024-07-08

  article1st authorCorresponding

  Artisanal supply chains play a significant role in the economy and social sustainability of many developing countries. The total value of the global artisanal and handicrafts market was $526 Billion (USD) in 2017 and is expected to reach $984 Billion (USD) by 2023 [NEST, 2018]. The sector is also the second biggest source of employment for rural women, especially for those from the under-privileged class. Unfortunately, while the artisan sector plays a crucial role, many workers in this sector struggle with low productivity and poverty, in part due to the highly distributed nature of the supply chain [Banik, 2017]. A significant fraction of artisanal production in developing countries is still conducted by artisans from their individual households that are dispersed across large geographical areas, thus posing numerous challenges including low productivity, limited access to financing, technology and quality control in the upstream of such supply chains.

  PublisherDOI

• Analytics and Operations to Improve Welfare in Distributed Artisanal Supply Chains

  Springer series in supply chain management · 2024-01-01

  book-chapter1st authorCorresponding
  PublisherDOI

• Design of Resale Platforms

  SSRN Electronic Journal · 2024-01-01

  articleOpen access
  PublisherDOI

• A Data-driven Approach to Improve Artisans' Productivity in Distributed Supply Chains

  SSRN Electronic Journal · 2023-01-01 · 2 citations

  articleOpen access1st authorCorresponding
  PublisherDOI

Frequent coauthors

• Georgia Perakis

  20 shared

• Omar Skali Lami

  IIT@MIT

  12 shared

• Leann Thayaparan

  Massachusetts Institute of Technology

  8 shared

• Qi‐Jun Hong

  Arizona State University

  7 shared

• Somya Singhvi

  6 shared

• Shelby Wilson

  5 shared

• Rachel J. Oidtman

  Merck & Co., Inc., Rahway, NJ, USA (United States)

  5 shared

• Simon I Hay

  5 shared