- Managed and updated modular dataset wtih annual online traffic of ~1.1 million users, displaying accessible information about proper disposal for 830+ garbage items.
- Increased the number of data points by 18% and streamlined data handling, reducing outstanding weekly change requests by 72%
- Assisted in website management for the city of Markham
- Improved marketing outreach by filming, editing, and deploying both video and static social media posts to improve community knowledge of waste management

This home security system uses opencv and haar classifiers to compare live video feed with reference images. Reference images are uploaded to the website. A python script does an api pull to get the reference images and constantly compares them to a live video feed, triggering a qnx based raspberry pi system to activate a water gun if a non-familiar face is detected.

This traffic data visualization tool was used to assist a proposal for reducing congestion in downtown Toronto. This model processed nearly 300,000 rows of csv historical data to identify congestion trends at peak times during morning, lunch, and evening rush hours. Using folium, it plotted the historical data as a heat map, making high traffic volume patterns clear. Key patterns were identified, leading to a strong pitch and solution involving congestion fees which ultimately led to a top 3 placement at the industry 4.0 competition.

Garbage Classification Tool Written as a side project in my position as a Waste and Environmental Student. Waste classification is a big problem. A lot of people have no idea where to put their garbage and can contaminate streams with waste that doesn't belong there. An example is putting a plastic bottle inside your composting bin. This causes difficulties for everyone on the product chain of this waste. The what goes where tool takes image input and outputs where you should put your garbage. Inputs are uploaded through the html frontend and are passed through openai gpt4o mini API, returning the disposal method.