ABOUT
I fell in love with data science in my high school statistics class, when I finally felt like I could use numbers to understand the world, tell real stories, and create positive change in the world. At UC Berkeley, I honed my technical expertise and deepened my understanding of the societal context behind data-driven decision-making. Armed with these skills, I stepped into the professional world, where I now turn complex data into actionable insights that drive impactful decisions and foster innovation.
A proud product of the Bay Area's innovation-centered mindset, I grew up in Silicon Valley, studied in the East Bay, and now call San Francisco home. Beyond work, I find joy in playing ultimate frisbee, writing sketch comedy, hiking in the great outdoors, and traveling. I’m always eager to tackle new challenges, whether that’s uncovering insights in a dataset, sprinting after a disc, or discovering my next favorite trail.
WORK EXPERIENCE
-
Conduct in-depth data analysis on gasoline consumption and EV transition using Python, PySpark, geospatial analysis, and the creation of Power BI dashboards
-
Primary analyst on multi-trillion row model of every US passenger car trip in the US
-
Author on 7 data-backed research reports as featured in the New York Times, Forbes, CBS News
-
Created data visualizations and factsheets for all 50 states and Washington DC, leading to direct policy influence in 18 states and more local governments
-
Connection point between internal data engineering team and public- facing policy/media team
Feb 2022 - May 2022, UC Berkeley
Researcher, Instanpolis Project
-
Parsed and visualized over 1,500 records of historical and ancient language census data in Python using NLP
-
Assisted in geospatial recreations of late Ottoman Istanbul/Constantinople
-
Discovered 8 new company verticals with by using NLP and machine learning on a topic-based model of customer text
-
Answered various company queries with the data-based tools Mixpanel, Looker, and Google Cloud/BigQuery
-
Created data pipelines to incorporate 5 publicly available international economic indicators into marketing research
EDUCATION
University of California, Berkeley (2019-2023)
B.A. Data Science, B.A. Political Science
Distinction in General Scholarship
Relevant Coursework:
-
Machine Learning
-
Data Mining and Analytics
-
Artificial Intelligence
-
Probability
-
Inference and Decisions
-
Computational Structures
-
Data Structures
-
Linear Algebra
-
Empirical Analysis and Quantitative Methods
-
Social Networks
SKILLS
Technical Skills
-
Proficient in Python, PySpark, R, SQL
-
Standard data science toolkit: pandas,scikit-learn, Matplotlib, various machine learning algorithms
-
Natural language processing, geospatial analysis, classical statistical analysis
-
Dashboard tools like PowerBI and Looker
Additional Skills
-
Public speaking
-
Strong writing ability
-
Collaborative team member, effective communicator
-
Experienced in policy data analytics and community engagement
LEADERSHIP &
VOLUNTEERING
Oct 2023 - Present, San Francisco
Event Organizer at Moishe House
-
Organize 7+ monthly cultural and social events to foster Jewish community
-
Plan, budget, and staff volunteers for events ranging from cultural celebrations to professional networking and volunteering
FEATURED PROJECTS
I co-authored and produced all data analysis and visuals for 2024's Cracking the Gasoline Code, a report by published by Coltura. This report used novel gasoline consumption data to inform how to cut gasoline use faster and more equitably with electric vehicles. It was featured in the New York Times and Forbes.
Cracking the Gasoline Code
/
Coltura
Comprehensive analysis introduces data-backed approaches to cut gasoline consumption, offering relief for those most impacted by fuel costs and promoting an accelerated transition to EVs.
I co-authored and produced all data analysis and visuals for 2024's Cracking the Gasoline Code, a report by published by Coltura. This report used novel gasoline consumption data to inform how to cut gasoline use faster and more equitably with electric vehicles. It was featured in the New York Times and Forbes.
I designed a comprehensive tool to evaluate NBA player value by predicting salaries using historical performance data and advanced statistics for my coursework in UC Berkeley's DATA 144. The project developed a two-model pipeline: one to estimate a player's fair salary based on key metrics and another to classify players into value categories like "elite" or "replacement level" based on their contributions. The tool aids teams in identifying undervalued talent, optimizing roster construction, and making informed salary decisions. Tested on a decade of NBA data (2011–2021), the tool demonstrated practical use cases for small-market teams and general managers by uncovering potential talent and providing insights into player market efficiency.
NBA Player Value Estimation Tool
/
UC Berkeley
Data-driven tool to assess NBA player value, predict salaries based on performance metrics, and classify players by their contribution level, assisting teams in roster optimization.
I designed a comprehensive tool to evaluate NBA player value by predicting salaries using historical performance data and advanced statistics for my coursework in UC Berkeley's DATA 144. The project developed a two-model pipeline: one to estimate a player's fair salary based on key metrics and another to classify players into value categories like "elite" or "replacement level" based on their contributions. The tool aids teams in identifying undervalued talent, optimizing roster construction, and making informed salary decisions. Tested on a decade of NBA data (2011–2021), the tool demonstrated practical use cases for small-market teams and general managers by uncovering potential talent and providing insights into player market efficiency.
Spearheaded the “Others” project during my internship at Promo.com, focused on understanding and classifying user behavior for a dataset of 44,000 users who labeled themselves as “Other” during onboarding. Leveraged Natural Language Processing (NLP) to process and cluster user-generated text data, identifying five distinct user verticals and classifying nearly 20,000 users. This project revealed critical insights into untapped customer segments and highlighted opportunities to align product offerings with diverse user needs. The work involved advanced Python skills, data cleaning, feature engineering, and visualization, showcasing my ability to handle complex datasets and derive actionable business insights.
NLP-Powered User Classification
/
Promo.com
NLP analysis to classify and analyze behavior across a dataset of 44,000 users, uncovering key customer insights.
Spearheaded the “Others” project during my internship at Promo.com, focused on understanding and classifying user behavior for a dataset of 44,000 users who labeled themselves as “Other” during onboarding. Leveraged Natural Language Processing (NLP) to process and cluster user-generated text data, identifying five distinct user verticals and classifying nearly 20,000 users. This project revealed critical insights into untapped customer segments and highlighted opportunities to align product offerings with diverse user needs. The work involved advanced Python skills, data cleaning, feature engineering, and visualization, showcasing my ability to handle complex datasets and derive actionable business insights.