Hi! I'm an assistant professor of Operations Research and Information Engineering (ORIE) at Cornell Tech, as part of the Jacobs Institute, and an ORIE, Computer Science, and Information Science field member at Cornell University.

Our work: "Full Stack, Public Interest AI"

I design, build, deploy, and evaluate public interest AI systems. Methodologically, my work spans computer science, operations research, machine learning/AI, and their intersection with economics and policymaking. I believe in having first-hand practical experience in a domain before tackling research questions, and work closely with several government agencies. As detailed below, my recent work is in two high-level directions:
  • Building and deploying public interest computational systems in collaboration with practitioners
  • Conceptually, mathematically, and empirically studying sociotechnical systems
See here for a recent "manifesto" on the challenges caused by heterogeneous participation in participatory systems, which also surveys my work broadly. See here for a recent talk video. Our work has received several awards, including the NSF CAREER, INFORMS George Dantzig Dissertation award, ACM SIGecom Dissertation Award (Honorable Mention), Forbes 30 under 30 for Science, the NSF graduate research fellowship, and several best paper awards. It has been supported by the Sloan Foundation, NSF, NASA, the Cornell Tech Urban Tech Hub, Google, Meta, and Amazon. Full bio.

Deployed Projects & Real-World Applications

We are building, deploying, and evaluating public interest computational systems. This work is often in collaboration with government agencies or non-profits using real data. Methodologically, we tackle challenges such as missing data, recommendations in equilibrium, and fairness in algorithmic decision-making. High school applications (with NYC Department of Education). We are studying disparities in how students apply to high schools in New York City, as a result of a complex process. For the 2025 cycle, we worked with NYC to help students in the application process. Bluesky Algorithmic feeds. We are building feeds on BlueSky for the academic community and beyond! Platform to help place discharged hospital patients into nursing homes. In Hawai`i, a PhD advisee built and deployed a platform to help place discharged hospital patients into one of over 1000 nursing homes, which are often run by single individuals out of their homes. The platform texts homes to ask for updated capacity and preference information, and then provides this information to about 10 hospital social workers. Resident crowdsourcing (with the NYC Department of Parks and Recreation). Understanding resident crowdsourcing behavior and government inspection and work-order resource allocation: do some neighborhoods report more than others for the same ground-truth conditions, thus receiving better government services? This has involved methodological work, as well as building and transferring a data dashboard and tree planting scheduling optimization. See here for a recent talk video. Library operations (with the New York Public Library) Do differences in how neighborhoods use the library hold system lead to all books "flowing" to a few neighborhoods, thus reducing access for others? We have quantified such disparities and are working on deploying optimized book flow procedures to mitigate them.

Theoretical Modeling, Methods Development, & Empirical Analysis

Our deployment work is informed by theoretical (mathematical) modeling, algorithm development, and empirical analysis. See here for a recent talk video, which also overviews my work generally. Understanding algorithmic monoculture How do algorithms make correlated decisions and errors, and what are the downstream implications in hiring (matching) markets, LLM-as-judge setups, and other applications? Fair, diverse recommendations in high-stakes settings How do we design recommendations systems in high-stakes settings that are fair (to both users and providers), and can provide diverse recommendations to users, preventing content rabbit holes? Computational Social Science more broadly How can we use modern NLP and optimization techniques to answer social science questions?

Contact me at ngarg@REMOVETHIScornell.REMOVETHISedu. Applicants: please read the information at the Contact page before emailing me.