Hello! I'm Cadey Chen, a fourth year Computer Science student at U Waterloo :)
Ever since a young age, I've been interested in applying historical data to understand our complex world. During my internships, I got to build a range of software that enable users to solve complex problems with their data. Whether it is meticulously analyzing simulation results with regression analysis; observing our eye and brain health through real-time eye tracking; or finding the root causes of recurring outages within a complex infrastructure.
Built and deployed new features for a data platform + analytics report design software. Enabled users to apply various statistical models on their simulation data and extrapolate key patterns; utilized Objected-programming to architect a new template class that enables users to generate a customizable multi-layered comparison report with an intuitive input table UI. Collaborated with users, product managers and designers to improve solution across iterations.
Developed new data processing pipelines, backend APIs, and frontend components for a cloud data platform that enables users to track and analyze the millions of eye data captured by their smart glasses at real-time. Leveraged concurrency, asynchronous programming, microservices, and message queues to efficiently process the data streams.
Improved the accuracy and performance of the outage root-cause analysis software by building new REST API methods that reliably scales to more concurrent users and implementing connection pooling for the MySQL database. Spearheaded the integration of Splunk and Avamar monitoring from Big Panda to Moogsoft and created alerting to appropriate support teams for different severity levels.
I trained a Machine Learning model that can analyze the genre and overall feeling of a movie using data from TMDB! On top of this ML model, I built a Full-Stack web application that gives users authentic movie recommendations when they are looking for something to watch. Users can choose to briefly describe the kind of movie they are looking for, or a similar movie they already love, or browse the top picks from each genre.
Python
Pandas
scikit-learn
Django
PostgreSQL
Bootstrap
I'm currently working with six other teammates from Wat.ai on DelayNoMore - a Machine Learning model with the goal of accurately predicting transit system delays in the city of Toronto given the time, date and transit route. We are currently using Supervised Learning on historical GTFS (General Transit Feed Specification) transit schedule and delay data to learn the delay patterns on each route.
python
pandas
scikit-learn
Wind and Sea is a Philosophy learning community where users can learn philosophy ideas from other users and also contribute to the knowledge database themselves by creating an account. For the Backend, I created a MongoDB database and built a REST API with TypeScript. The API supports user creation, user authentication, as well as data retrieval, insertion, update, and deletion by authenticated users.
TypeScript
JavaScript
MongoDB
A responsive and interactive web application where users can learn and share knowledge about philosophy in a vibrant online knowledge community. Users can create an account, browse resources in the community, and add new resources to the community. I designed the UI from scratch on Figma, the overall style is inspired by classical art.
React
Redux
CSS
Figma
An interactive single-player game built with Object-oriented Programming principles using C++. The player can choose their character from different mystical legends; and then travel through five randomly generated dungeons with hidden treasures, mysterious potions and gruesome enemies. The player need to defeat attacking enemies, and collect as much treasures as possible so they can safely travel to the top dungeon. Due to university policies, the code cannot be published online.
C++
During my academic exchange, I took DH-411: Design Research for Digital Innovation at the EPFL + ECAL Lab. Our team of four applied design research theories and conducted user tests in iterations to come up with the optimal app design. The final product - Nomad is a healthy lifestyle mobile app that impacts users to develop healthier habits by emphasizing community building and recommending fun healthy activities in their own neighborhood based on their schedule.
Figma