Sumi-e koi swimming toward an enso
Stanford ’30

jason shi

Mathematics / Computer Science

Portrait of Jason Shi at Stanford

Hi! I'm Jason, a first-year at Stanford studying mathematics with an interest in CS and ML. I like learning about new things, working on hard problems, and making things that look nice and help other people.

Recent hobby projects have included creating data pipelines that read tens of thousands of job postings a day, backtesting models from public data sets, and game-theoretic code strategies. Before Stanford, I co-ran Ladue Math club, whose videos have reached 50M+ views.

Studying
Mathematics & Computer Science, Stanford ’30
Classes
MATH 61CM · STATS 116X · CS 106B
Honors
US Presidential Scholars Semifinalist
Tsinghua International Summer School
4× AIME qualifier
Elsewhere
GitHub · LinkedIn · Ladue Math Club Instagram

Selected projects

  1. Bridge 渡 HTML · CSS · JS
    Summer 2026     Director's Choice Award at Tsinghua. A bilingual palliative care navigation app scoped as a WeChat Mini Program, led with a team of five.
  2. Internship Opportunity Tracker Python · GitHub Actions · Azure OpenAI
    Sept. 2026 —     Reads about 33,000 postings a day from 250 sources and opens a ranked digest as a GitHub Issue each morning. Tracks 571 open internships missing from the most-used public list.
  3. Random Forest Portfolio Backtester Python · scikit-learn · pandas
    Summer 2026     A from-scratch backtester that holds one of 15 stocks or cash each day. Returned +250% out of sample (2021–26) before costs, retested on shifted windows to check for overfitting.
  4. Citadel Terminal Strategy Algos Python · forward simulation
    Summer 2026     11th of 30 finalist teams worldwide. A frame-by-frame combat simulator scores every attack and defense plan within the 5-second turn limit.

Stanford University Mathematics Camp · Algebraic Topology · July 2025

Topological Games and the Baire Category Theorem

Jason Shi and Kevin Xiao · Mentored by Andrew Lee · 14 pages

Abstract

Alice picks an open interval of ℝ, Bob picks one inside it, and they keep taking turns forever. Bob wins if every point the intervals share is rational. Alice has a winning strategy, because ℚ is “small” in ℝ. We build up the theory of games like this one, namely the Choquet, strong Choquet, and Banach–Mazur games, and use them to characterize Baire spaces and prove the Baire Category Theorem. We then apply it to the structure of real-valued functions.

Theorem (Oxtoby). A topological space X is Baire if and only if Alice does not have a winning strategy in the Choquet game on X.

Application. No function f : ℝ → ℝ is continuous at only the rationals.

Read the paper ↗ Repository ↗

Say hello!

GitHub · LinkedIn

Or leave a note