A deterministic rollback-vs-naive netcode desync benchmark — proof that localhost-only netcode desyncs under latency/loss — with an interactive in-browser demo.

A 3D Slicer code benchmark that mechanically catches the hallucinated and deprecated Slicer/VTK/ITK APIs LLMs emit, graded against a curated symbol catalog, with a NiiVue medical-image viewer.

A paraview.simple API-version benchmark: statically check whether a pvpython script's APIs exist in a given ParaView version, against a curated migration table.

A benchmark for the one thing LLMs can't do in 3D — wire a node graph. Validates Geometry-Nodes graphs (sockets/types/cycles) against a real Blender catalog, with a React Flow viewer.

FreeCAD edit-resilience benchmark: sweep a driving dimension and measure whether downstream features survive — robust vs brittle parametric modeling, with a 3D dimension-slider dashboard.

MLT-XML compiler with a render-in-the-loop oracle: turn an edit spec into a melt-renderable project that provably loads and renders frame-exact — plus the dataset of where LLMs break on timeline fi…

Spec-to-circuit benchmark: candidate circuits simulated in ngspice and graded against quantitative specs, with the naive-netlist traps LLMs fall into.

Reproducible GeoAI benchmark + eval harness (PostGIS/GDAL/GeoServer) where LLMs fail — verifier-graded tasks, dashboard, and training-data export.

CLI copytrading bot for Polymarket — mirrors target trader's BUY + SELL trades

A Spring Boot API for science teachers to convert between different units of measurement in their worksheets.

A Spring Boot UI application for science teachers to convert between different units of measurement in their worksheets.

This is a single-page application (SPA) built with React and Material-UI (MUI) that allows users to search for movies and bookmark them for later viewing. The application fetches movie data from a …

Created a personal website to display my resume, past projects and experiences, as well as my understanding of Web-development. Link: http://kevinwu.pythonanywhere.com/

A platform to promote entrepreneurship and collaboration amongst college students. It provides an efficient and fast way to quickly get their idea out there. Creates a way for them to quickly recru…