Eliot Seo Shekhtman.

I am a Ph.D. student in Computer and Information Science at the University of Pennsylvania, where I am advised by Professor Nikolai Matni. My research interests are in the intersection of machine learning, control theory, and learning theory for dynamic systems, with the goal of studying fairness in deployed systems.

I additionally have a CS MS from the College of Computing and Information Science at Cornell University with a minor in Linguistics, where I was advised by Sarah Dean. Previously, I worked as a contingent machine learning engineer for Goldman Sachs in the Compliance Engineering division.

I can be reached through email at shekhe at seas dot upenn dot edu, and I have public Google Scholar and LinkedIn profiles.

In my free time, I like cooking!

Penn Engineering | Aug 2024-present | Philadelphia, PA

Ph.D. in Computer and Information Science

Advisor: Nikolai Matni

Bowers CIS | Aug 2022-May 2024 | Ithaca, NY | [ thesis ][ slides ]

Masters of Science in Computer Science with a minor in Linguistics

Advisor: Sarah Dean

College of Engineering | Aug 2018-May 2022 | Ithaca, NY

Bachelors of Science in Computer Science with a minor in Linguistics

Magna Cum Laude with Honors

Sep 2013-Jun 2018 | Delmar, NY

Aug 2024-present | Philadelphia, PA

Research direction: system identification of dynamic systems

• Researched information-theoretic parameter recovery bounds on dynamic systems.

Jan 2023-May 2024 | Ithaca, NY

Research direction: learning theory for dynamic systems

• Researched regret bounds of strategic classification in a contextual bandit setting.
• Characterizing convergence of multi-learner systems in the presence of strategic behavior by users.
• Proved convergence conditions for a strategic usage setting, and published single-author work to AISTATS 2024.

Sept 2022-Apr 2023 | Ithaca, NY

Research direction: natural language processing

• Researched and implemented applications of latent diffusion models to text generation.
• Published to NeurIPS 2023.

Advisor: Madeleine Udell | Jun 2020-Jul 2022 | Ithaca, NY

Research direction: unsupervised learning

• Published a machine learning paper in AAAI-22 on the imputation of missing data using a Gaussian copula model.
• Implemented frameworks and benchmarks to study active learning for distribution shift.

Advisor: Marlen Gonzalez | Jun 2019-Dec 2021 | Ithaca, NY

Research direction: computational neuroscience

• Cleaned and scored data and automated administrative tasks in the lab using R, improving in data analysis skills.
• Wrote Python scripts under supervisor Dr Gonzalez to implement the Balloon Analog Risk Task (BART) for fMRI and for mTurk.
• Processed participants and collected data using the BART, Stroop, and Cyberball tasks in the Cornell fMRI scanner for later analysis.

Advisor: Igor Lednev | Jun 2016-Aug 2017 | Albany, NY

Research direction: computational biology

• Wrote MATLAB scripts under supervisor Dr Lednev to conduct Singular Value Decomposition (SVD) analysis on 1GB of cell images for the classification of cellular patterns characteristic of lead toxicity.
• Prepared buffer solutions, grew both bacterial cells and human kidney cells, and tested/observed the effects of lead and EDTA on the latter, using light microscopy.

Role: Machine Learning Engineer @ Compliance Engineering

Sept 2022-Mar 2025 | Remote

• Developed and productionized an internal distributed automated machine learning (AutoML) package capable of training and evaluating thousands of models simultaneously.
• Consulted to improve and fix existing algorithms and training methods.
• Productionized an LLM training pipeline, utilizing test-driven development.

Role: Engineering Summer Analyst @ Compliance Engineering

Jun 2022-Aug 2022 | New York City, NY

• Trained BERT-based transformer models to classify text documents, achieving high recall and good workload reduction on a test set.
• Developed a semi-supervised model to highlight sections of text influential for a classification task.
• Implemented a generalized pipeline for training large NLP models for future projects.

Role: Engineering Summer Analyst @ Compliance Engineering

Jun 2021-Aug 2021 | New York City, NY

• Implemented MapReduce jobs to efficiently filter, tokenize, label, and embed ~10 million articles stored in Google Protocol Buffer format.
• Researched and developed LSTM and attention models to predict on a Named Entity Recognition task, achieving high accuracy over ~5 million sentences.
• Established a pipeline for training and evaluating future NLP models efficiently on new data, and constructed an API leveraging the model for ease of access.

Role: Graduate Teaching Assistant

Spring 2024 | Ithaca, NY

• Created programming questions and exams.
• Created reference materials for the class elaborating on lecture material.

Role: Head Teaching Assistant

Fall 2023 | New York City, NY

• Created homework assignments, programming questions, quizzes, and exams.
• Set grading deadlines and ensured everything was running according to schedule.
• Organized regular meetings with teaching assistants and graders.
• Held homework recitations to review student questions on the solutions.

Role: (Graduate) Teaching Assistant

Spr 2021-Spr 2023 | Ithaca, NY

• Created and graded assignments and exams.
• Held in-person office hours.

Role: Teaching Assistant

Fall 2020 | Ithaca, NY

• Held a recitation and created unique powerpoints every week expanding on both current material and on specific concerns raised by students.
• Held online office hours and graded assignments.

Role: Instructor

Jun 2019-Sep 2020 | Ithaca, NY

• Taught Python, HTML, CSS, and block-based languages to students from first to eleventh grade, developing interpersonal communication skills.
• Developed a curriculum for teaching Python and adapted in-person curriculums for online learning.

Role: Tutor

Sep 2017-May 2018 | Delmar, NY

• Taught high school math to other students in preparation for the NYS Regents exam.
• Prepared students for Algebra 1, Algebra 2, Geometry, and Precalculus.

May 2022-Jan 2023 | Ithaca, NY

• Helped with computer science and linguistics consultation in a project to create a functional online dictionary for the Gayogohó꞉nǫʔ language.
• Researched Gayogohó꞉nǫʔ morphology and developed software to parse nominal structures into their constituent parts, enhancing the searchability of morphologically-complex words.

Feb 2020-Jun 2022 | Ithaca, NY

• Developed an online algorithm to predict room usages and optimize energy consumption by the HVAC system in Upson Hall.
• Coordinated with two other subteams in 20+ meetings to gather information.
• Led the development of a script to generate accurate dummy data for initial algorithm training.
• Algorithm is forecasted to reduce unnecessary HVAC operation by over 70%.

Feb 2021 | Ithaca, NY

• Developed a novel analytical approach in a group to approximate the final fraction of adopted nodes for a contagion in a clique based network.
• Implemented simulations to test the approximation, demonstrating consistency and accuracy of the model, and wrote a research paper on the results.