Open to new-grad roles & part-time work with early-stage teams

Bryan Chen

Statistics, Economics & CS student at the University of Toronto, currently in the data org at Wattpad. I build machine learning systems that answer questions I actually care about — from projecting MLB performance to reading market sentiment. I take data projects from raw source to deployed product (pipeline, model, infrastructure, and interface) which is the same thing an early team needs to start up its data and analytics from scratch.

View my work Get in touch GitHub

At a glance

Now: Engineering Intern, Wattpad; Data org
Studying: Statistics, Economics & Computer Science; University of Toronto
Focused on: Machine learning & data systems; Plus a lot of computer vision
Based in: Toronto, Canada; Open to relocating

01About

Building at the intersection of ML and the things I love

I'm a Statistics, Economics & Computer Science student at the University of Toronto with a focus on machine learning and data systems. I love building things — especially the side projects that uncover random patterns in the intersections of my interests in sports, music, and ML.

Right now I'm wrapping up my engineering internship in the data org at Wattpad, and simultaneously building a computer vision tool to analyze tennis swing mechanics. The goal: make quality coaching feedback accessible to anyone with a phone.

Always looking for interesting problems and people to learn from. Whether it's ML, tennis, or house music, please reach out.

Machine learning

Clustering, fine-tuning, and the evaluation work that decides whether a model actually ships.

Data systems

Pipelines, lakes, and the infrastructure that keeps those models fed on a schedule.

Computer vision

Pose estimation and video analysis — currently pointed at a tennis court.

02Work

Selected projects

Things I built end to end — model, pipeline, infrastructure, and the interface on top.

FeaturedFormerly Diamond-DNA

xWAR-Engine

Projecting MLB performance, one season ahead

A full-stack projection system that forecasts next-season batting and pitching lines — PA, wOBA, AVG/OBP/SLG, HR and SB for hitters; IP, ERA, FIP, WHIP and K/BB rates for pitchers. A comparables engine pairs GMM archetype clustering with kNN similarity to surface each player's closest historical analogues. Everything runs on a bronze/silver/gold S3 data lake with scheduled Lambda pipelines, provisioned end to end in Terraform.

PythonScikit-LearnGMMkNNAWS LambdaStep FunctionsS3TerraformReactTypeScriptVite

Visit live site Source

FinSense

Market sentiment that teaches itself

A fine-tuned FinBERT classifier scores news about a given equity as negative, neutral, or positive, surfaced as a per-symbol heatmap. Confident predictions become new training data while uncertain ones route to an LLM for labeling, so the corpus grows daily with no manual annotation — then a SageMaker pipeline retrains on it and registers the new version only if it clears a macro-F1 gate.

PythonPyTorchFinBERTLLMsAWS SageMakerDynamoDBTerraformGitHub ActionsReact

Visit live site Source

In progress

Tennis swing analysis

A computer vision tool that breaks down swing mechanics from phone video, so quality coaching feedback doesn't require a coach. Currently working through pose estimation and building the frame-by-frame comparison view.

PythonComputer VisionPose Estimation

03Skills

What I work with

The stack I reach for when taking an idea from notebook to production.

Languages

PythonRSQLTypeScriptJavaScript

ML & Data

PyTorchScikit-LearnPandasNumPyFinBERTComputer VisionGMMkNN

Cloud & Infrastructure

AWS LambdaAWS SageMakerS3API GatewayDynamoDBStep FunctionsEventBridgeTerraformdbt

Tools & Frameworks

GitGitHub ActionsReactViteNext.jsNode.jsTailwind CSS

04Experience

Where I've worked

Engineering Intern
Current
Sept 2025 — Present
Wattpad
Build and maintain dbt and Airflow pipelines that transform billions of rows of event data into the analytics tables the product team runs on.
Research Assistant
Current
Sept 2023 — Present
University of Toronto
Designed and deployed real-world experiments with the French department, then analyzed the resulting linguistic data.
Data Analyst Intern
May 2024 — Aug 2024
University of Toronto Scarborough
Analyzed institutional data in Python and Power BI to deliver insights that raised student internship hiring rates by 20%.

05Interests

Outside of work

The things my side projects tend to grow out of.

Sports

Tennis · Hockey · Baseball

I play tennis as much as I can — it's where I release my competitive energy. Avid Canucks fan (honestly I'd love for any Canadian team to win a cup) and Blue Jays fan when I'm not on the court.

Music

DJing · Production · House

I produce and DJ mostly electronic music: deep house, tech house, drum and bass, and everything in between. There's a lot of overlap between engineering a set and engineering software.

Building

Side Projects · ML · Computer Vision

Building random things is how I learn best. If I'm curious about something, I'll build a project around it — hence the tennis swing analysis tool.