Jason Li

Expected May 2027

B.S. Computer Science

University of Wisconsin–Madison · GPA 4.0 / 4.0

Relevant coursework Artificial Intelligence · Learning Theory · Linear Algebra · Theory of Probability · Introduction to Stochastic Processes · Linear Optimization · Algorithms · Operating Systems · Software Engineering · Object-Oriented Programming and Data Structures

May 2026 – Present

Software Development Intern

Epic

• Working on the frontend design and implementation of Hyperspace, Epic's primary application used by healthcare organizations.

Nov 2025 – Jun 2026

Research Assistant

University of Wisconsin–Madison

• Characterized the Nash Equilibrium for a multi-player "battling influencers" game.
• Extended the analysis to the Bayesian (incomplete-information) version of the game and characterized its Bayesian Nash Equilibrium (BNE).
• Wrote the mathematical proofs and contributed to drafting the paper.
• Wrote Python programs to compute and visualize game behaviors across different scenarios.
• This research contributes to a paper to be submitted to GameSec 2026.

Jan 2026 – May 2026

Peer Mentor

CS 540: Introduction to Artificial Intelligence · UW–Madison

• Ran weekly office hour sessions, 5 hours per week, answering students' questions and helping them work through their homework.
• Taught core AI and machine learning concepts, such as lasso regression.

May 2025 – Aug 2025

Software Engineering Intern

Bohr Systems

• Designed a camera-only visual localization pipeline in Python for quad-wing UAVs in GPS-denied environments; defined system interfaces and data flows.
• Compared visual-localization methods on requirements and trade-offs; recommended ORB (adopted).
• Performed a thorough code-level review of the OpenCV ORB implementation, covering all major algorithmic steps and performance optimizations.
• Implemented an end-to-end ORB localization prototype: keypoint extraction, descriptor matching, geometric verification (RANSAC), and map-tile search.
• Built Gazebo simulation experiments; scripted camera streams and test harnesses to evaluate accuracy.

May 2024 – May 2025

Research Assistant

Beijing Academy of Agriculture and Forestry Sciences

• Assisted in developing deep learning algorithms for sturgeon re-identification (ReID) in Python & PyTorch.
• Reviewed fish detection and individual identification (ReID) methods.
• Acquired, processed, and curated a training and validation dataset of 6,000 high-quality sturgeon images, designing the image-acquisition protocol and labeling guidelines.
• Fine-tuned and integrated YOLOv8 as the detection module in the pipeline; generated detector-aligned crops for ReID training/inference.
• Adapted a pose-guided ReID approach to sturgeon.

Feb 2026 – May 2026

Online Multiplayer Boggle Game

CS 506 team project · React · Spring Boot · MySQL · UW–Madison

• An online, multiplayer version of the board game Boggle, with matchmaking lobbies, timed rounds, and user accounts with score tracking.
• Built the multiplayer gameplay on the backend in Java and Spring Boot, including joining and managing game lobbies, running live timed rounds, server-side events that keep each player's game state consistent, and tallying players' final scores.
• Set up the backend foundation that stores user accounts and scores, runs the server in a container, and connects the web app to the backend.
• Wrote automated JUnit tests, including MVC tests, run through a CI/CD pipeline, helping the team reach over 80% code coverage.
• Implemented frontend features like the game over results screen, a lobby full pop-up window, and the game rules page.
• Acted as product owner, setting the team's coding standards and presenting the sprint backlog.

Jan 2025 – May 2025

Partially Observable Markov Game Solver

Directed study with Dr. Young Wu · UW–Madison

• Implemented Nash Q-learning and deep Q-learning in Python and PyTorch to compute Nash equilibria in partially observable Markov games.
• Analyzed algorithmic solvability across game scenarios; identified solvable cases and diagnosed causes of non-convergence.

Jan 2025 – May 2025

Intersection Detection for Autonomous Driving

SAE-sponsored competition

• Led vehicle-to-intersection distance estimation for the team.
• Reviewed existing detection methods, training workflows, and pre-trained models, evaluating performance and computational requirements.
• Implemented OpenCV filters, edge detection, and line detection in Python to estimate the intersection location.