Ryan Kim
Incoming CS @ UIUC — building ML systems and full-stack tools.
I’m Ryan Kim — a senior at William Fremd High School and incoming Computer Science student at UIUC (Fall 2026). I build ML systems from scratch, systems-level software, and open-source tools.
Current focus: clinical ML research at Memorial Sloan Kettering and Weill Cornell, plus non-coding DNA prognostic modeling as a Simons Fellow at Stony Brook.
Quick links: Projects · Publications · CV page
Experience click any row to expand
▸ 
Jun 2025 — Present
DNABERT-based prognostic model for glioblastoma survival via Cox regression on non-coding regulatory DNA mutations; identified age-related survival signals and racial disparities encoded in mutational scores.
▸ 
Lead Technical Developer · MathLinks.org Nov 2025 — Present
Raised $20k in funding (Hudson River Trading, Innovate901 First Place). Built a full-stack math competition platform serving 150k+ members across 11 partner organizations.
▸ 
Research Intern · Advanced Computing & Oncology Lab, Memorial Sloan Kettering Dec 2024 — Jan 2026
Built a DSPy-optimized LLM pipeline to extract disease progression events from CT/PET/clinical reports. Co-authored abstract on referral event analysis accepted at ACRO 2026.
▸ 
Research Intern · AI in Medicine & Computational Biology Lab, Weill Cornell Medicine Oct 2024 — Present
Investigated systematic bias in AI tumor purity estimators across TCGA and internal cohorts; demonstrated that scaling data volume and model size fails to resolve generalization gaps across cohorts.
▸ 
Founder · Papers2Code.org Jan 2024 — Present
Founded platform indexing 300k+ ML/AI papers lacking public code implementations; built pipeline to map papers to GitHub repositories.
Show older experience ▾
▸ 
Full-Stack Developer · Cyberlinc, Inc. Aug 2023 — Jul 2024
Engineered a full-stack crowdfunding platform (Flask / SQLAlchemy) with secure payment integration.
Selected Projects All projects →
microDINOv3
2026Complete DINOv3 self-supervised ViT training system in dependency-free pure Python — student-teacher EMA, multi-crop augmentation, centering mechanism.
Papers2Code
2024 — PresentPlatform indexing 300k+ ML/AI papers without public code. Scraping and mapping pipeline connecting papers to GitHub.
Deep MoE Reproduction
2024Open-source reproduction of "Deep Mixture of Experts via Shallow Embedding" — a paper that had no public implementation.
MathLinks.org
2025 — PresentFull-stack math competition platform for 150k+ members across 11 partner organizations. $20k raised.
Publications All publications →
CPU-only LLM pipeline (LLaMA3.1-8B, Qwen3-4B with DSPy / MIPROv2) to classify lung cancer CT reports by specialty referral category; analyzed 370 clinically annotated reports at MSK.
Skills
Python 
Java 
C/C++ 
SQL 
JavaScript 
HTML/CSS 
PyTorch 
TensorFlow 
Pandas 
NumPy 
Scikit-learn 
Matplotlib 
Plotly 
R 
Polars 
React 
Next.js 
Express 
Flask 
FastAPI 
MongoDB 
SQLAlchemy 
Git 
Linux 
Docker 
VS Code