Elizabeth Santorella
Research Engineer at Meta
Medford, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolElizabeth Santorella is a Research Engineer at Meta with a decade of experience blending economics, statistics, and engineering to build practical ML and experimentation infrastructure. Trained as an economist (MA, PhD) with an SB in Physics and Economics from MIT, she moved into data science and scientific software, owning end-to-end pricing and causal-inference systems that drove billions in revenue. At Meta she focuses on adaptive experimentation and has contributed to high-profile open-source projects like Facebook's Ax and PyTorch/BoTorch, improving model interfaces, scheduler callbacks, and CI-docstring quality for Bayesian optimization tools. Her strengths are statistical computing, causal inference, and experiment design, coupled with hands-on backend and DevOps work that keeps research code production-ready. Comfortable across Python, C/C++, Spark, and Docker, she also brings mentorship and hiring experience from small-team environments. A detail many miss: she combines rigorous academic research with practical software fixes—like Pareto-frontier and type-checking improvements—that materially improve optimization tooling used by practitioners.
10 years of coding experience
7 years of employment as a software developer
SB, Physics, Economics, SB, Physics, Economics at Massachusetts Institute of TechnologyMaster's (MA), Economics, Master's (MA), Economics at Harvard University
English, Spanish
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (16)
documentation10
model-building10
pytorch10
machine-learning10
bayesian10
flake10
optimisation10
python10
modeling10
model-driven10
model-driven-development10
optimization10
testing10
continuous-integration9
api8
Programming languages (5)
Github contributions (5)
facebook/Ax
Jul 2022 - Jan 2023
Adaptive Experimentation Platform
Role in this project:
Backend Developer
Backend DeveloperContributions:23 reviews, 15 commits, 308 PRs in 5 months
Contributions summary:Elizabeth's contributions primarily involve modifying the `ax` library, focusing on the `models` and `modelbridge` submodules. They removed data requirements for TorchModels before candidate generation and raised a `DataRequiredError` in the `BotorchModel.fit` function. The user also added callback functionality to the Ax Scheduler and included an example of how this can be used for in-place figure updates during the optimization process. Additionally, the user addressed issues with the Pareto frontier computation, corrected type signatures and other fixes related to outcome constraints for MOO and other type-checking improvements.
experimentationadaptivesimulationadaptive-experimentation-platformexperimentation-platform
meta-pytorch/botorch
Jul 2022 - Jan 2023
Bayesian optimization in PyTorch
Role in this project:Backend Developer & DevOps Engineer
Backend Developer & DevOps EngineerContributions:3 releases, 273 reviews, 36 commits in 5 months
Contributions summary:Elizabeth contributed to the project by fixing docstrings and adding checks for them in the continuous integration workflow using flake8-docstrings. They updated the project documentation by expanding the docstrings and creating new documentation for models. The user also fixed failing tests related to long docstring lines, and optimized the optimization and error handling of functions called during model training and acquisition function optimization.
bayesianpytorch
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.