Chicago, IL • Reed College '23, Mathematics & Statistics
After graduating from Reed College with a degree in Mathematics and Statistics, I made a deliberate choice: before building tools for learners, I wanted to understand them. 700+ tutoring sessions later, I have a clear picture of where and why people get stuck in math. Most students do not have a broad math problem; they have a gap in a specific foundational concept that was never caught, and once that gap is identified, progress is measurable and fast.
That instinct for what actually moves a student got rigorous at Yuno Learning, where the job was to prove whether an intervention worked rather than assume it did — A/B tests on content cadence and deadlines (t-tests, Cohen's d) and conversion tracking from a first website visit through to a tutor match. I learned to trust measurement over intuition.
Current focus: an IBEW aptitude test prep platform (launching 2026), where I get to do both at once — building my own product and instrumenting it end to end. The core feature is a free diagnostic that tells candidates exactly where they stand, what to study, and how long they need. Behind it is an analytics-heavy backend that turns raw performance data into personalized guidance, an LLM-generated coach summary, and full-funnel tracking so I can measure which acquisition channels actually convert.
Technical toolkit: Python, R, SQL, statistical modeling, experimentation design, A/B testing, and funnel/event instrumentation.
Outside of that, I spend my time powerlifting, following automotive design, and checking out cool places around Chicago for a chill work session. I started chicagoworkspots.com to collect work-friendly spots, initially for me and my partner. One simple but important feature is a power outlet rating based on the often-overlooked seat-to-outlet ratio. Feel free to check it out and add your own suggestions!
A summer 2022 internship that continued as intermittent freelance analytics work through August 2025. Queried and reshaped tutoring cohort data in SQL for weekly trend snapshots, authored A/B test readouts (t-tests, Cohen's d) on content cadence and deadline expectations, and helped build conversion-tracking dashboards spanning first website visit through tutor match.
Coach 7 students weekly across K-7 Math, SAT Math, and AP Calculus/Physics; recently took one student's SAT Math score from 580 to 730 in three months. Built and run scotchegg.co, a free 1,440-question SAT Math practice platform that doubles as an acquisition funnel for tutoring clients.
Delivered 600+ one-on-one sessions across K-12 through college-level math and statistics, and designed introductory Statistics and Probability curricula for non-traditional learners.
Conducted applied statistical analysis on higher-education outcomes data, including work that identified a significant discrepancy in public earnings estimates and supported institutional reporting discussions.
An end-to-end diagnostic platform for electrical-trades apprenticeship candidates. Candidates take a free practice test and get a breakdown of their strengths and gaps, a prioritized study plan, and a realistic timeline to readiness. Built around an adaptive assessment engine, a student-facing analytics dashboard, and an LLM-generated coach summary (Claude Haiku). I am also instrumenting a full-funnel acquisition test across three channels (TikTok, forums, Facebook) to measure which one converts best.
Curated a 1,236-school ESSA dataset and built random forest models in R (caret, randomForest, SMOTE, 9-fold cross-validation) to predict ELA/Math proficiency, reaching over 80% accuracy for most cohorts and surfacing staffing metrics as the top predictive drivers. GitHub ↗
Analyzed a large-scale advertising experiment to evaluate the impact of ad exposure on conversion, using hypothesis testing and confidence intervals to separate real lift from noise, then weighed conversion efficiency against cost to land on a clear go/no-go recommendation. GitHub ↗
A free SAT Math practice platform (React, Supabase) with 1,440+ questions organized across ten topic units, built to double as an acquisition funnel for my tutoring. Ships with a browsable catalog, downloadable worksheets, and build-time SEO prerendering. GitHub ↗
A prototype that uses a vision LLM to generate progressive, one-step-at-a-time hints for SAT Math problems from a photo — scaffolding a student toward the answer instead of handing it over. GitHub ↗
A normalized PostgreSQL schema for a trading-card domain — modeling players, packs, rarity rules, and pack openings with a focus on correctness and clean, queryable structure. GitHub ↗