PeopleofSuperintelligence

Researcher

Nathan Kallus

Nathan Kallus is an operations research researcher whose work connects machine learning with causal inference and personalized decision-making. He earned a mathematics B.A. at Berkeley in 2009 and completed an operations research Ph.D. at MIT in 2015. From 2015 to 2016, he served as a postdoctoral associate in MIT's Operations Research and Statistics group and as a visiting scholar in USC's Department of Data Sciences and Operations before joining Cornell as an assistant professor for 2016–17. His work developed methods for personalizing treatment choices and evaluating decision policies from observational data; his coauthored 2025 paper on experiment rules won KDD's Best Paper Award in Applied Data Science. As of October 2026, he is an associate professor of operations research and information engineering at Cornell Tech and Cornell Engineering and Director of Machine Learning & Inference Research at Netflix.

Timeline

  1. 2009–2015

    Berkeley degrees and an MIT PhD

    At UC Berkeley, he earns a B.A. in mathematics in 2009. He also earns a B.S. in computer science there. He completes a Ph.D. in Operations Research at MIT in 2015.

  2. 2013

    An MIT student paper award

    In 2013, he and Vishal Gupta win the MIT Operations Research Center Best Student Paper Award for “Robust SAA” and become finalists in the INFORMS Nicholson Best Student Paper Competition for “Data-Driven Robust Optimization.”

  3. 2015–2016

    MIT postdoc and USC visiting scholar

    From 2015 to 2016, he serves as a postdoctoral associate in MIT's Operations Research and Statistics group at the Sloan School of Management and as a visiting scholar in USC's Department of Data Sciences and Operations at the Marshall School of Business.

  4. 2016

    Recursive treatment personalization

    In 2016, he proposes recursive partitioning to choose the treatment with the best causal effect for each person, using observational data to create regimes where different treatments are optimal.

  5. 2016–2017

    Assistant professor at Cornell

    For 2016–17, he serves as an assistant professor in Cornell's School of Operations Research and Information Engineering.

  6. 2017

    Balanced policy evaluation

    In 2017, he proposes a balance-based method for evaluating and learning personalized decision policies from observational data. The method finds weights that balance historical data with a target policy.

  7. 2025

    A KDD applied science award

    In 2025, he coauthors “Evaluating Decision Rules Across Many Weak Experiments,” which uses cross-validation to assess rules for choosing treatments in A/B tests. KDD awards the paper its Best Paper Award in Applied Data Science.

  8. As of October 2026

    Associate professor and Netflix director

    In July 2026, he coauthors oci-agent, an open-source workflow with a human in the loop for observational causal inference. In August 2026, he coauthors a Netflix experiment study with 8.5 million users that measures how recommendation improvements change the products they consume. As of October 2026, he is an associate professor of operations research and information engineering at Cornell Tech and Cornell Engineering and Director of Machine Learning & Inference Research at Netflix; he is a coauthor of Applied Causal Inference Powered by ML and AI, a book on combining causal inference with machine learning.

Career

  1. NowCornell UniversityAssociate Professor of Operations Research and Information Engineering, Cornell Tech and Cornell Engineering
  2. NowNetflixDirector of Machine Learning and Inference Research
  3. 2015–2016MIT Sloan School of ManagementPostdoctoral Associate
  4. 2015–2016University of Southern California Marshall School of BusinessVisiting Scholar
  5. Cornell UniversityAssistant Professor, School of Operations Research and Information Engineering, Cornell Engineering

Education

  1. 2015Massachusetts Institute of TechnologyPhD in Operations Research
  2. 2009University of California, BerkeleyBA in Mathematics
  3. University of California, BerkeleyBS in Computer Science

From the biography.

Checked career sources: tech.cornell.edu, news.cornell.edu, news.cornell.edu, events.stanford.edu, nabe.com, arxiv.org, arxiv.org, nathankallus.com.

Sources and credits

Photo: We have no published photo of Nathan Kallus yet. Send us a free-licensed photo.

  1. Cornell Tech - Nathan Kallus tech.cornell.edu. Checked 27 October 2026.
  2. Introducing new members of the faculty for 2016-17 | Cornell Chronicle news.cornell.edu. Checked 27 October 2026.
  3. MIT ORC - Seminars - Spring 2014 web.mit.edu. Checked 27 October 2026.
  4. Recursive Partitioning for Personalization using Observational Data arxiv.org. Checked 27 October 2026.
  5. Balanced Policy Evaluation and Learning arxiv.org. Checked 27 October 2026.
  6. Awards – KDD 2025 kdd.org. Checked 27 October 2026.
  7. Evaluating Decision Rules Across Many Weak Experiments arxiv.org. Checked 27 October 2026.
  8. A Human-Augmenting Agentic Workflow for Observational Causal Inference arxiv.org. Checked 27 October 2026.
  9. Recommendation Quality and the Concentration of Consumption: Experimental Evidence from Netflix arxiv.org. Checked 27 October 2026.
  10. Nathan Kallus nathankallus.com. Checked 27 October 2026.