If you’re looking for my one-page resume, look no further. A more detailed description of my experience is below.
Work Experience
Software Engineer Intern at Meta - Facebook AI Applied Research
I worked on the AI for Ads team on the Ads Supply project. Our goal was to model the impact of changing ad load on revenue and engagement metrics at the user level. I developed and applied variants of meta learner models for causal inference and analyzed these models on a variety of performance metrics and with white-box model explainability techniques. I also proposed and implemented new techniques for more accurate model evaluation.
Data Science Intern at Verizon - Systems Performance Engineering
I worked with systems performance data to analyze and optimize network parameters affecting dropped call KPIs. I developed pipelines to automate data aggregation and trained a variety of machine learning models including XGBoost ensembles to select optimal parameter values for network configuration.
Teaching Assistant for 15-150: Principles of Functional Programming
As a teaching assistant, I co-taught a lab/recitation section with over 20 students. developed new content for homework assignments. helped over 100 students at office hours, and graded assignments and exams.
Grader for Art of Problem Solving
As a grader, I provided personalized feedback to student problem submissions for Python and algebra problems.
Research Experience
Evaluating Synthetic Code-Switched Data
As part of the 07-300 and 07-400 research practicum sequence, I worked on generating and evaluating synthetic multilingual code-switched text for use in finetuning code-switched language models. This project was documented through this website at Code-Switching.
Verdant: Computational Notebook Versioning
As part of the 2020 CMU HCII REU program, I worked on the JupyterLab extension Verdant, which provides advanced version control features for experiments run in computational notebooks, including cell-level versioning, output diffing, and intelligent search.
I worked on development of the core extension with React, Redux, and TypeScript. I also developed a chart classification pipeline using scikit-learn to differentiate different types of chart output images from other images.
AI-Assisted User Research Tools
As part of a CMU HCII independent study, I developed a Python tool for automatically summarizing research interview data and performing comparisons between textual data to find outliers. I also performed a landscape analysis of current methods in abstractive and extractive text summarization.
Architecture and Analysis for High-Assurance Autonomy
I worked on a project to improve training and testing methods for computer vision neural networks, using self-driving planes as the problem domain. I coauthored a paper on input prioritization methods for lowering the human cost of labeling collected data. I coauthored another paper on generating synthetic test data using variational autoencoders. I also wrote interfaces for and automated data collection in the X-Plane 11 simulator.
Projects
LU-Partition
We implemented the LU-partition algorithm from this paper as part of a project with the Princeton Gerrymandering Project’s Ensemble Club. The algorithm will be used in algorithmic drawing of district maps, with the goal of improving map analysis through direct sampling of the map distribution.
QA-QG System
Our Natural Language Processing semester project was a question-answering and question generation system. Given a set of Wikipedia articles, it can either generate factoid questions about the text or answer given questions. Our project won the position for best question generation.
This project was built primarily with spaCy and NLTK, with some components such as BERT models from Hugging Face.
Recipe Delta
Recipe Delta was a HackCMU 2019 project. We created a web application that allows you to easily find recipes oriented towards using up ingredients you have on hand.
Our frontend was built with AngularJS and our backend was built with Python.
EduPass
I worked on this project from my freshman through junior years of high school, first as part of the JA Company program and then as part of an independent startup.
We created an educational technology product to solve the problem of tracking school attendance during flexible schedule offerings (i.e. periods in which students were free to choose what classes or extracurricular offerings they participated in). Our tool allowed students to search offerings and coordinate with friends, teachers to invite students and track classroom attendance, and administrators to view individual to school-level analytics across multiple sessions.
I held a variety of positions throughout my time with EduPass, including frontend developer (in AngularJS), backend developer (LAMP stack with Python and Flask), and CTO (managing product direction and coordinating adoption of Agile management strategies).
Coursework
Carnegie Mellon University
GPA: 4.33/4.00
Spring 2023
- 10-605: Machine Learning with Large Datasets
- 10-613: Machine Learning Ethics and Society
- 10-708: Probabilistic Graphical Models
Carnegie Mellon University
GPA: 4.00/4.00
Fall 2022
- 11-324: Human Language for Artificial Intelligence
- 15-330: Introduction to Computer Security
- 15-784: Cooperative AI
- 85-213: Human Information Processing and Artificial Intelligence
Spring 2022
- 07-400: Research Practicum in Computer Science
- 10-707: Advanced Deep Learning
- 15-418: Parallel Computer Architecture and Programming
- 76-221: Books You Should Have Read By Now
Fall 2021
- 05-318: Human AI Interaction
- 07-300: Research and Innovation in Computer Science
- 10-725: Convex Optimization
- 11-492: Speech Processing
Spring 2021
- 10-315: Introduction to Machine Learning (SCS Majors)
- 11-488: Computational Forensics and AI
- 15-317: Constructive Logic
- 15-451: Algorithm Design and Analysis
- 84-352: Representation and Redistricting
Fall 2020
- 11-411: Natural Language Processing
- 15-210: Parallel and Sequential Data Structures and Algorithms
- 16-385: Computer Vision
- 33-104: Experimental Physics
- 36-401: Modern Regression
- 98-163: Student Taught Courses (StuCo): Introduction to Tetris
- 98-317: Student Taught Courses (StuCo): Hype for Types
- 98-380: Student Taught Courses (StuCo): Practical Economics
Summer 2020
- 15-213: Introduction to Computer Systems
Spring 2020
- 05-589: Independent Study in HCI-UG
- 07-180: Concepts in Artificial Intelligence
- 15-150: Principles of Functional Programming
- 15-251: Great Ideas in Theoretical Computer Science
- 15-281: Artificial Intelligence: Representation and Problem Solving
- 36-226: Introduction to Statistical Inference
Fall 2019
- 07-128: First Year Immigration Course
- 07-131: Great Practical Ideas for Computer Scientists
- 15-051: Discrete Math Primer
- 15-122: Principles of Imperative Computation
- 15-151: Mathematical Foundations for Computer Science
- 16-161: ROB Freshman Seminar: Artificial Intelligence and Humanity
- 21-241: Matrices and Linear Transformations
- 76-101: Interpretation and Argument
- 99-101: Computing @ Carnegie Mellon
University of Minnesota
GPA: 4.00/4.00
The University of Minnesota’s PSEO program allows dual enrollment in UMN classes during high school. I enrolled in this program full-time during 12th grade (2018-2019), and took the following courses:
Spring 2019
- CSCI 1933: Introduction to Algorithms and Data Structures
- MATH 5248: Cryptology and Number Theory
- MATH 5652: Introduction to Stochastic Processes
- MATH 5707: Graph Theory and Non-Enumerative Combinatorics
Fall 2018
- ENGL 3027W: The Essay
- MATH 5535: Dynamical Systems and Chaos
- MATH 5651: Basic Theory of Probability and Statistics
- PHAR 1002: Medical Terminology
- POL 3308: Congressional Politics and Institutions
University of Minnesota Talented Youth Math Program (abbreviated UMTYMP): The collegiate component consisted of the courses MATH 1471, MATH 1472, MATH 1473, MATH 2471, MATH 2472, and MATH 2473, covering single and multivariable calculus and linear algebra. I took these courses from 2014 to 2017, spanning 8th through 10th grade.
Volunteering
CMU Science Olympiad
As part of CMU SciOly, I wrote tests for the Fossils and Codebusters events. I also proctored and graded exams for the Fossils event for multiple years.
CMIMC
I contributed to the development of the website used for CMIMC remote competitions during the pandemic, working with Django and JavaScript.
Free Geek Twin Cities
I worked with a team of volunteers to build computers out of donated parts and install Linux on these devices for discounted sale to low-income families. I also took part in the disassembly and recycling of old electronics.