Emma Castiglia
Research Data Scientist at Meta
Yale Particle Physics Ph.D., 2022
Netflix ML Research Intern, 2021.
Emma Castiglia
About Me
Data Scientist. AI Enthusiast. Physicist. Foodie. Runner.
I am a Research Data Scientist at Meta working in the privacy program. I am applyiing and evaluating ML models to improve the way we maintain our privacy commitments to our users, as well as creating and monitoring business metrics for the program. I received my PhD in physics from Yale where my research focused on applying machine learning to improve detector performance and optimize signal sensitivity at the ATLAS experiment. I recieved my BA in both physics and mathematics from the University of Chicago. To improve my ML knowledge before moving to industry, I took Deep Learning courses at Yale, attended ML focused summer schools, and completed an ML research internship.
Outside of physics and ML, I am deeply committed to improving gender equity in STEM. At Meta, I have particiapted in the Women in Data Lean-in circles, Women in Tech mentoring program, and the IDS Allyship circle. I served as the chair (co-chair) for 3 years for both Women in Physics+ and Equity in the Job Search Symposium at Yale and have advocated for diversity and inclusion initiatives within my department. I really enjoy teaching and inspiring younger students and taught for 12 semesters at Yale and was a part of 3 women in STEM mentoring programs.
Meta: Infrastructure Data Science - Privacy Program
Resarch Data Scientist
Since August 2022, I have been a research data scientist at Meta within the Infrastructure Data Science Org partnering with the Privacy Program. I established a data-driven framework for the commitments team with clear North Star metrics to goal towards each half as well as consolidated all data tooling into one easy-to-find place. In my first half at the company, I led the Tech Audit & Privacy data science, tooling, and AI teams to create a single, collaboratitve roadmap to prioritize projects for H1 2023. I also prototyped a pipeline to apply ML to search for privacy commitments in external documents to comply with FTC assessor requests. Finally, I designed a study to evaluate privacy ML models performance and their ability to create business impact.
ATLAS Experiment
ML Applications to Particle Physics
For 8 years I researched for the ATLAS collaboration at CERN. During graduate school, I worked on a search for a rare Higgs boson decay to tau leptons. I trained a neural network to separate out the main background that most closely resembled our signal. The neural network score was the final discriminating variable in our fit model. I also worked on ML improvements for the Tau Trigger by optimizing a BRT to improve energy calibration and explored using an equivariant neural network in the Tau ID. I spent a year working on a Higgs boson to ZZ_dark search using a BDT, which combined the buzz wordy topics of ML, Higgs boson, and dark matter!
ML for Systems at Netflix
ML Research Intern, Summer 2021
I was lucky enough to be a part of Netflix's first ever summer intern class where I spent 12 weeks working with the Algorithms Engineering and Compute Teams. I studied the feasibility of using time series predictions to model CPU requests to their cluster. This could allow for predictive autoscaling on the cluster to make it run smoother and more efficiently, while reducing wait times for employees and users.
Mu2e Experiment at Fermilab
Research at the Intensity Frontier
During my second year at Yale, I conducted research for Mu2e which is a planned Intensity Frontier Experiment at Fermilab. It is trying to detect the charged lepton flavor violating decay (which would be a beyond the Standard Model process) of a muon decaying directly to an electron. I worked on calorimeter triggers for both signal electrons and high energy photons and benchmarked other MVA algorithms in the trigger system.
Yale University
PhD Student, 2016-2022
I had the great joy of completing my thesis under the direction of Professor Sarah Demers at Yale. I explored many of Yale's course offerings and took classes in Deep Learning, where I trained a CNN to classify famous paintings. I also served as a Teaching Fellow for 12 semesters in courses ranging from Graduate Particle Physics to Introductory Physics Labs and the Physics of Music. I also served as the Chair of the Graduate Women in Physics+ group and ran the Yale Physics Professional Development Organization Seminar series.
University of Chicago
Undergraduate Student, 2012-2016
I attended the University of Chicago, where I double majored in physics and math, studied abroad in Paris (with the Paris Math Program), and filled my free elective spaces with art classes. I spent my junior and senior years, plus the summers before, researching for the ATLAS experiment under Professor Young-Kee Kim. This is where I learned how to code and got my start in particle physics research working on a Higgs boson coupling dark matter search and simulation studies for the Fast TracKer. The photo here is of when my physics classmates and I handed in our Honors theses!