Eytan Bakshy

I’m a Director and Research Scientist at Meta, where I lead the Adaptive Experimentation team within Central Applied Science. Our group develops methods and systems for efficient learning and optimization, spanning both foundational research and applied impact across the company. We work across many domains, including AI training and inference, conversational and generative systems, recommender systems, hardware design, and physical modeling. My research interests include Bayesian optimization, preference learning, multi-armed bandits, and active learning. I’m especially passionate about democratizing these methods through open-source software and shared research infrastructure. Our team develops BoTorch, a research framework for Bayesian optimization, and Ax, a platform for adaptive experimentation.


Selected publications

Here are a few of my recent publications. For a more complete list of publications and activities, see my Google Scholar page and likely outdated CV.

Bayesian optimization, active learning, and reinforcement learning

Causal Inference and Experimentation

Field experiments

Observational studies

Open source

The Adaptive Experimentation team at Meta maintains two open-source projects

I am also the creator of PlanOut. It's an open-source framework for designing and implementing complex behavioral science experiments. I find PlanOut to be helpful for thinking about and planning experiments, and hope that people can use it to run interesting experiments. Visit the PlanOut homepage and give it a try. For more on designing and deploying online field experiments with PlanOut, check out our paper on the subject.

Press

Coverage on papers and Data Science posts from the early days: Other press