About Me

I'm a student, software engineer, freelancer, and data science enthusiast.

At school, I'm a fourth year Computer Science Co-op (Hons.) student at the University of Manitoba, specializing in Software Engineering and Artificial Intelligence.

At work, I'm a software engineer and business analyst with experience in web and mobile app development, data analytics, and robotic process automation (RPA). In addition to that, I also do some freelance app development.

At home, I like to experiment with and learn about data science and machine learning, put together the occasional electronics project, and work on various software projects (such as this website).

My non-tech interests include music (I play guitar), personal finance, physical sciences, and the outdoors.

Projects

Poker Timer (source) — An Angular-based webapp to time the blinds for a poker tournament.

Traffic Accident Data Explorer (source) (paper)

Fitness App — A Flutter IOS app built for a freelance job, featuring a Firebase Backend.

Productive (source) — An neatly-engineering Android app for tracking tasks, and collecting virtual items for their completion.

Meal Ordering Site UI (source) — A made-up meal ordering service similar to Skip or DoorDash, with a total focus on UX and design.

GAN Experiment (source) (paper)

This Site (source) — A site to showcase my projects, built using three.js.

Traffic Accident Data Explorer

Using a data mining technique, known as multidimensional multilevel quantitative association rule mining on a public dataset of UK Traffic incidents from 1979-2020, we (a group of 4) generated over 27,000 association rules. The implementation of the multidimensional multilevel quantitative association rule mining algorithm was done in Python using Numpy. Separately, I created web-based explorer tool using Angular to help us search through the association rules and find ones that we thought were interesting or insightful. Through this, we were able to locate some interesting results. For example, in our data set, of all the night time accidents that are fatal, 87% of the time the driver was male.

The associated paper was awarded Best Paper in the 2022 IEEE IRI Conference, and will be published in the next edition of the conference proceedings.

Fitness App

Built as part of a freelancing job, the IOS fitness plan app allows users to view various fitness plans, track their progress with their account and watch tutorial videos as they complete the exercises. It was built using Flutter for the frontend and Firebase in the backend. Release to the app store is coming shortly.

Productive

Productive is essentially a gamified task tracker, kind of like Trello. It was built in a group of 5 for a software engineering course, thus, clean code and project organization was at the forefront of this project. We utilized a 4-tier architecture for the project; I managed most of the backend. The Android-based app used an ORM library to remove the need for database boilerplate code, and included a large number of unit, system, and integration tests, resulting in about 85% code coverage of the project.

An Experiement with GANs

This research paper covers the relatively new neural network variation known as a GAN. It introduces the concept of a GAN in a somewhat simple manner, compares it to a classic neural net, and outlines an experiment that was performed in order to test different variations of training data.
I edited an existing GAN implementation so that the training ratio of positive to negative samples could be varied. This gave me my first real experience working with PyTorch, and enriched my machine learning knowledge as a whole. It was originally written for a course project, and was not submitted for publication.