Co-Founder / CTO

• Doing cool things on the computer. Hopefully the last job I'll ever have. :')

---

Machine Learning Researcher

• Worked with a team of researchers building a legal foundation model aligned with human preferences trained on data from court rulings, laws, and regulations.
• Built pipeline for labeling legal text sentence-level rewards for conditional human preference pretraining as in Pretraining Language Models with Human Preferences.
• Owned evaluation pipeline for benchmarking state-of-the-art language models (including Claude, GPT-4, and our own models) on legal tasks like bar exam questions, CaseHold, and LegalBench.

---

Founding Engineer

• Privacy and legal startup using technology to facilitate trust and promote access to justice. Scoped, architected, and built full-stack web application with Next.js and Tailwind. Drove key decisions related to product design and strategy.

---

Empirical Research Fellow

• Led project with the IRS to improve audit selection with machine learning, and worked with Santa Clara County on data-driven interventions using active learning to measure and improve COVID-19 monitoring and control.
• Developed algorithms for probabilistic race and ethnicity imputation, offline geocoding, and bootstrapped phylogenetic analysis to analyze the efficacy of COVID-19 interventions and their effects on health equity.
• Applied recent self-supervised tabular deep learning models to improve data efficiency for predicting tax noncompliance.

---

Software Engineer

• Built a cloud-based data analysis platform to analyze battery sensor data. Wrote scripts to simulate IoT devices, and built out a secure data pipeline to securely stream telemetry data to the cloud.
• Used cloud software such as AWS IoT Core, AWS Kinesis streams, Elasticsearch, and Kibana dashboards to transform, analyze, and visualize telemetry data.

---

Data Engineer

• Selected for Bluebonnet Fellowship, which trains technical graduate and undergraduate students to work with political data, and matches them with political campaigns as volunteers.
• Led efforts to build data infrastructure for the Christy for Congress campaign, including automated data transformation from SQL database into reports and tables in Google Sheets.
• Built a dashboard web application using Dash to automatically generate reports on key campaign performance indicators.

---

Investment Associate Intern

• Produced a research report on threats posed to Amazon by antitrust, and implications for the stock market.
• Wrote 5+ case studies of sovereign defaults; analyzed causal mechanisms to develop theory of debt crises.
• Conducted qualitative research and exploratory data analysis on global metals markets.

---

Energy Analyst

• Developed and tested statistical models to forecast natural gas fundamentals, and researched the future trajectory of renewable energy, focusing on the impact of tax credits and energy prices.
• Underwent rigorous hands-on training in trading energy markets, with a focus on power and natural gas.
• Participated in seminars with traders and business school professors, covering topics such as valuation, behavioral finance, and how to trade markets including equities, commodities, and currencies.

Leveraging Genomic Sequencing Data to Evaluate Disease Surveillance Strategies

We use whole-genome sequencing data to measure the effectiveness of different COVID-19 disease surveillance strategies used by Santa Clara County to identify possible transmission links.
[Journal Link] [PDF]

---

A Language-Matching Model to Improve Equity and Efficiency of COVID-19 Contact Tracing

✨ Recognized with the Innovative Practice Gold Award by the National Association of County and City Health Officials

We implement an interpretable language-matching model to predict likelihood of patients being low English-proficiency Spanish speakers, in order to assign them to Spanish-speaking contact tracers and avoid the friction of translation services.
[Journal Link] [PDF]

---

Evaluation of Allocation Schemes of COVID-19 Testing Resources in a Community-Based Door-to-Door Testing Program

We use an active learning approach based on the COVID-19 positivity rate to sample locations to send door-to-door Spanish-speaking health workers in vulnerable communities. We compare this approach to allowing the community health workers to leverage local knowledge.
[Journal Link] [PDF]

---

(Not) Your Type: Race, Dating, and Wrongful Discrimination

✨ Winner of the Lyle and Olive Cook Prize for Best Honors Thesis

I analyze theories of wrongful discrimination, and explore how they might be applied to the problem of racial preferences in romantic and sexual relationships. I lay out the ethical problems associated with racial preferences, and survey the existing social science research. I introduce two philosophical accounts of what makes discrimination wrong—one based on demeaning, and another based on harm. I consider what resources these theories of discrimination can provide in service of an argument about the moral status of racial preferences. Finally, I consider the limitations of viewing racial preferences through the lens of discrimination.
[PDF]

Stanford University

M.S. in Computer Science, 4.0 GPA

• Coursework: Parametric and non-parametric statistics, machine learning theory, natural language processing, computer vision, reinforcement learning, Bayesian networks, big-data mining, algorithm design and analysis, computer systems, computer and network security, cryptography, complexity and information theory, linear algebra, multivariate calculus, web applications.
• Teaching: Led discussion & developed exams for the course “CS142: Web Applications,” covering HTML/CSS, JavaScript, React, databases, network security, & frameworks for building large-scale production web applications.
• Academic Distinctions: Selected for Siebel Scholars Program, a fellowship awarded to a handful of outstanding graduate students studying computer science, business, and bioengineering in the top graduate programs in the country.

---

Stanford University

B.A. in Philosophy with distinction, 4.1 GPA

• Coursework: Logic, metalogic, ethics, epistemology, metaphysics, philosophy of science, feminist philosophy, political philosophy.
• Academic Distinctions: Phi Beta Kappa, interdisciplinary honors in Ethics in Society, Lyle and Olive Cook Prize for top Ethics in Society honors thesis, Boothe Prize for Excellence in First-Year Writing.

• Programming Languages: Proficient in Python (including PyTorch), JavaScript/Typescript (front-end and back-end), and R (tidyverse and bioinformatics). Limited knowledge of C++, Rust, and Golang.
• Other Skills: Training, debugging, and evaluating machine learning models in PyTorch, distributed training, Weights & Biases, HuggingFace Accelerate and Transformers libraries, linear algebra, Docker, AWS, FastAPI & Express, React, Next.js, Tailwind, numpy, pandas, Jupyter, SQL, databases, static site generators.

• Stanford Debate Society (President 2018–19; Captain 2016–17): Led a debate team ranked top-five in the nation, which is completely self-funded and supports domestic and international travel and competition expenses for around 50 students every year.
• Responsibilities: Helped to hire and manage 6 employees; led tryouts for and trained potential new debaters; managed short- and long-term financial planning and budgeting; managed staff and handled logistics for large debate events.
• Competitive Achievements: 2nd Place at 2019 North American Championship; top-32 partnership at 2018 and 2019 World Championships; 18th-ranked partnership at 2017 World Championships.
• Hobbies and Interests: International politics, classical music, baking artisanal bread, stand-up comedy.