Jason Poulos
Human Frontier Collective Specialist at Scale AI
Boston, Massachusetts, United States
Summary
👤
Senior🎓
Top SchoolJason Poulos is an independent researcher and machine learning specialist with 11 years of experience at the intersection of ML, causal inference, and policy, currently designing domain-specific evaluations for frontier AI at Scale AI. He holds a Ph.D. in Political Science with a Designated Emphasis in Computational Science and Engineering from UC Berkeley and completed postdoctoral work applying transformers and causal methods in clinical and observational health settings at Brigham and Women’s Hospital and Harvard Medical School. His research portfolio includes multi-valued treatment causal methods, Bayesian frameworks for adversarial ML, and deep-learning approaches to missing-data imputation and counterfactual prediction, bridging methodological rigor with practical evaluation design. Based in Boston, he translates safety concerns into measurable tests for large models and trains ML engineer agents by creating benchmark solutions, combining academia-grade research with hands-on benchmark engineering. An often-overlooked strength is his ability to port causal thinking into domain-specific evaluation tasks, making complex policy-relevant questions machine-testable.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD Political Science with a Designated Emphasis in Computational Science and Engineering, Doctor of Philosophy - PhD Political Science with a Designated Emphasis in Computational Science and Engineering at University of California, BerkeleyBachelor of Arts - BA Economics, Bachelor of Arts - BA Economics at University of Massachusetts Amherst
Github Skills (53)
missing-data10
representation-learning10
pytorch10
causal10
causal-discovery10
feature-selection10
lime10
ocr10
machine-learning10
imputation10
causality10
inference10
causal-inference10
statsmodels10
deep-learning9
Programming languages (3)
Github contributions (5)
jvpoulos/jvpoulos.github.com
Dec 2016 - Jan 2025
Contributions:164 pushes in 8 years 2 months
jvpoulos/MCPanel
Sep 2019 - Dec 2021
Matrix Completion Methods for Causal Panel Data Models
Contributions:51 commits, 48 pushes in 2 years 4 months
principal-component-analysispanel-datafactorizationcausallinear-models
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.