HOBBIES

OUTSIDE
THE CODE

What I do when I'm not building. Running long distances, playing poker, and competing in volleyball.

RUNNING

I run long distance and run out of ideas. I run for fun. Minus the fun. Twelve Half marathons, 30k runs, not chasing the next PR just building habits. Currently goal is Ironman 70.3 2028.

VIEW ON STRAVA ›

— All-Time km

— km (365 Days)

— Runs (365 Days)

— Time (365 Days)

RUNNING LOG

MonWedFri

RUN

SWIM

RIDE

PERSONAL BESTS

Syncing from Strava…

RECENT RUNS

Syncing from Strava…

HIGHLIGHTS

Tamarack Marathon — May 2026 — 43.16 km

Road to Recovery — Sep 2025 — 21.11 km

THAT'S THE HARD PART

Last synced from Strava API: —

POKER

Competitive tournament poker player at the University of Waterloo Poker Club. Study the system, find the edge, execute under pressure. Convincing myself that poker is a skill game and pretending I have an edge. Spending hours of studying to fix my leaks and increase my edge by a half a percent.

PLACEHOLDER

2 / 891 Fall Season Rank

33% Final Table Rate

6 / 18 Tournaments Final Tabled

~250 Avg Field Size

FALL 2025 SEASON

Finished 2nd out of 891 active members across the Fall 2025 UW Poker Club season — the largest student poker club in Canada. Final tabled 6 of the 18 tournaments I entered, putting me in the top 9 of an average field of 250 players per event. A 33% final table rate in fields that size isn't luck. It's sustained edge.

Following the season I joined the club as an executive for the Winter term, helping run games and grow the competitive structure of the club. Currently studying ICM pressure and 3-bet pot construction with the goal of making the UW IPA poker team.

AT THE TABLE

FINAL TABLE

875,000 chips. Starting stack: 5,000. The kind of run where every decision compounds.

VARIANCE

Running KK into AA on the last two tables of a midterm tournament. You can sharpen every edge, and still run face-first into variance. Learning to separate process from outcome is the real skill.

VOLLEYBALL

Competitive player. Photos and clips coming soon.

COMING SOON

Media and writeup in progress.

NELSON KANG

Education

University of Waterloo

Bachelor's Degree, Computer Science and Finance (CFM)

Coursework: Advanced Functional Programming (Racket), Data Structures & Algorithms (C), Linux/Bash/Git Tooling, Linear Algebra, Calculus
Activities: Poker Club Executive, Competitive Intramural Volleyball Captain

Experience

Ericsson Canada Inc. Jun 2024 – Aug 2024

Software Engineer Intern — Ottawa, ON

Developed 4 ETL pipelines using C# and SQL Server to extract Jira data from JSON payloads, transform custom field mappings, and load into reporting tables; used temporal tables for full change-history tracking.
Audited and catalogued all existing team pipelines, identifying faulty custom-field mappings producing NULL data; created a reference spreadsheet that became the team's primary pipeline debugging resource.
Maintained Power BI dashboards for trouble report visibility, supporting stakeholder reporting and data-driven decision-making across the BI team.

Projects

ClipFarm — Automated Volleyball Highlight Generator Mar 2026 – Present

Python, FastAPI, Next.js, TypeScript, YOLOv8-pose, Roboflow, PostgreSQL, Docker, Cloudflare R2

Engineered a dual-detection video pipeline: YOLOv8-pose classifies player actions while a custom-trained Roboflow ball detector tracks trajectory to identify rally boundaries, reducing false-positive clips by ~40%.
Containerized with Docker Compose; FastAPI serves auth-gated endpoints, Celery + Redis distributes processing jobs, and Modal offloads GPU inference. Auth via Supabase JWKS with Row-Level Security.
Implemented the frontend in Next.js + TypeScript with a filterable clip grid, inline multi-label correction UI for future model training, clip trimming with background re-cut via Celery, and lazy-loaded thumbnails.

Multi-Factor Robo-Advisor — 1st Place, Market Beat Competition Nov 2025

Python, Pandas, NumPy, yfinance

Optimized a $1M CAD portfolio across U.S. and Canadian equities; achieved 5.7% return vs. 3.7% market average benchmark, placing 1st out of 18 teams.
Implemented a multi-factor portfolio construction pipeline applying momentum, volatility, correlation, and trend signals inspired by the Ranked Asset Allocation Model (RAAM).
Automated end-to-end: data ingestion, liquidity screening, factor ranking, sector-cap enforcement, and Sharpe-ratio optimization under real-world constraints (transaction fees, FX conversion, position limits).

ChessAI — Neural Network Chess Engine Nov 2025

Python, PyTorch, Modal

Trained a neural-network chess engine on millions of Stockfish-labeled positions using distributed GPU compute on Modal; designed a scalable ML pipeline for ingesting and labeling large Lichess datasets.
Integrated classical alpha-beta search with deep neural evaluation for real-time gameplay, balancing search depth with inference latency.

QuillifyEdu — Educational Content Management Platform 2024

C#, ASP.NET Core, SQL Server, Entity Framework Core, JavaScript

Developed a full-stack content management platform using ASP.NET Core 8.0 with Razor Pages, enabling users to create, organize, and share educational notes with a database-driven backend.
Implemented ASP.NET Core Identity for production-grade authentication including email confirmation, account lockout policies, and enforced password complexity requirements.

Technical Skills

LanguagesPython, C#, C, SQL, TypeScript, JavaScript, HTML/CSS

FrameworksFastAPI, Next.js, .NET, PyTorch, Pandas, NumPy, Celery, SQLAlchemy

InfrastructurePostgreSQL, SQL Server, SSMS, Supabase, Redis, Modal, Cloudflare R2, Git, Linux, Power BI

ToolsJira, FFmpeg, OpenCV, YOLOv8, Selenium, Alembic, Docker, Roboflow

DOWNLOAD PDF ›

OUTSIDETHE CODE