ryan@xing:~/portfolio
menu

$

Ryan Xing — CS @ Vanderbilt.

I build full-stack apps, AI workflows, dev tools, and mobile software. Seeking Summer '27 SWE internships.

$ cat experience.log --reverse-chron

May 2026 - Present

Undergraduate Researcher Algoverse AI Research Program

Remote
  • Latent failure prediction in long-horizon LLM agents — paper in preparation for NeurIPS.
  • Reproduced a published method for predicting whether an AI coding agent will succeed, matching the foundation paper's accuracy within 8%, by rebuilding its value-axis signal from scratch inside the Qwen3-8B model.
  • Extended the signal to long-horizon coding agents, a regime the original paper never evaluated, by engineering a Python pipeline that extracts model activations at every step of 60-step coding tasks.
  • Implemented statistical measures of how reliably a value-axis signal predicts success across LLM agent attempts at fixing real GitHub bugs from SWE-bench.

July 2026 - August 2026

Software Engineering Intern North York Chinese Baptist Church

Toronto, ON
  • Centralized announcements, events, programs, forms, and profiles into one place for 1,000+ church members by building a trilingual React Native and Expo app with Supabase authentication and PostgreSQL row-level security.
  • Enabled staff to publish content without engineering help by creating a trilingual pipeline of 3 serverless functions that syncs Google Sheets to 15 PostgreSQL tables daily.
  • Reduced repeat data entry by 30% by designing a configurable forms engine that validates submissions and autofills member details from church records.
North York Chinese Baptist Church

./nycbc.png

May 2026 - June 2026

Software Engineering Intern Canadian Youth Champions

Toronto, ON
  • Delivered a production survey platform (Next.js, FastAPI, PostgreSQL) for public campaigns that gathered an estimated 3,000-10,000 multilingual responses each, creating one centralized system to launch, moderate, and analyze campaigns.
  • Cut survey translation from 2 hours to minutes with an LLM workflow, unblocking new-language launches.
  • Replaced manual review of thousands of survey responses by developing LLM-powered analysis of sentiment, bias, and themes across multilingual campaigns.
  • Implemented a 3-stage GitHub Actions CI pipeline (pytest/Vitest tests, Docker builds, CodeQL security scans) that runs on every pull request to catch regressions before release.
Canadian Youth Champions

./cyc.png

June 2024 - January 2026

Full-Stack Developer Incendium Academy

Remote
  • Cut median page-load time from 2.5s to 1.5s by adding image compression, code splitting, and caching, keeping lessons usable for 300+ monthly students.
  • Lowered registration drop-off from 50% to 40% with funnel analytics, putting ~30 more students into free courses.
Incendium Academy

./incendium.png

$ ls featured/ --long

./recall --statsready

Hackathon

GenAI Genesis

Storage model

Local IndexedDB

Search mode

RAG + rerank

output / project metadata ● loaded

ReCall

featured build

Chrome extension that turns browser history into a semantic search and knowledge graph experience.

  • Delivered a GenAI Genesis project using RAG, embeddings, reranking, conversational search, and graph visualization over browsing history.
  • Used a browser extension so capture happens where browsing context already lives.
PlasmoReactTypeScriptTailwind CSSIndexedDBview project ↗
./health-protocol-tracker --statsready

Swift files

48

Repositories

7

Unit test suites

2

output / project metadata ● loaded

Health Protocol Tracker

featured build

SwiftUI iOS app for peptide inventory, injection logs, reminders, and pharmacokinetic concentration modeling.

  • Created a real-use iOS tool with vial inventory, blend support, dose validation, reconstitution flow, schedules, injection history, and half-life charts.
  • Kept Firestore access out of SwiftUI views by routing all persistence through seven domain repositories.
SwiftSwiftUIFirebaseCombineSwift Chartsview project ↗

$ cat skills.json

"languages"
Python · TypeScript · JavaScript · Java · SQL · C/C++ · Bash
"frameworks_libraries"
React · Next.js · Node.js · FastAPI · Express.js · Flask · React Native · tRPC · Expo · Tailwind CSS
"tools_infra"
Git · GitHub · PostgreSQL · Docker · GitHub Actions · pytest · Vitest · CodeQL · MongoDB · Vite · uv · Ruff · Google Colab
"ml_data"
PyTorch · scikit-learn · pandas · NumPy · LLM APIs · XGBoost · CatBoost · SHAP · Recharts · Jupyter
Ryan Xing

./headshot.png

$ cat about.md

I'm a computer science student at Vanderbilt who likes testing new ideas by turning them into real products. My work spans full-stack systems, AI research and workflows, developer tools, and mobile apps.

Outside of computer science, I'm into underground and electronic music, including Ninajirachi, Underscores, The Hellp, and Oklou. Some favorite stories are Avatar: The Last Airbender, Beef, Parasite, and Code Geass.

$ ps aux | grep ryan # currently

  • Researching latent failure prediction in long-horizon language agents.
  • Building trilingual member tools at North York Chinese Baptist Church.
  • Open to Summer 2027 software engineering internships.

$ mail -s "internship" ryan

Let's  build something.

ryanxing06@gmail.com
© 2026 Ryan Xing(づ。◕‿‿◕。)づ