Lumo is a weather forecasting platform that uses machine learning to predict weather patterns rather than the numerical models used in traditional forecasting methods. Lumo predicts precipitation and cloud activity for the near future by leveraging satellite imagery and atmospheric data. Generating a forecast in this manner is more efficient than current methods, shortening the run time of each prediction and enabling the use of the most recent weather data.

Nowadays, many people are too busy to prepare nutrition-balanced meals for their families. Commonly, families do grocery shopping once a week, and they do not know how to follow a healthy diet with the groceries they bought. EatSmart is a personalized and low-cost intelligent meal assistant aiming to plan healthy daily meals based on ingredient availability, users’ health conditions, and preferences while keeping a low food wastage. EatSmart consists of both mobile and voice assistant applications that support interactions through a natural language.

At Philadelphia Legal Assistance, 95% of mandated legal aid requests are rejected because the firm doesn’t have the resources to adequately take those cases. Our platform, LegalRelief, aims to help North Americans by crowdfunding fees for legal cases. A community of retail investors will fund these anonymized legal cases and settlements from those cases will be proportionally divided amongst the investors. The platform will facilitate the funding as well as the matching of legal cases to legal counsel.

When out with friends and family, we all take pictures on our phones. Oftentimes, everyone has a small subset of photos, and we’re left scrambling for the rest. We remove the step of having to ask for and upload pictures afterwards so everyone can have them instantly. Additionally, when sharing photos, there is often a cluttered feed of repeats. We alleviate these issues through a mobile app where users can take photos and automatically add them to a live collaborative album, while grouping similar photos.

Each co-op cycle, approximately 8000 students from the University of Waterloo attempt to find a co-op job through WaterlooWorks. Usually, students can only consider the job posting provided by the employer when applying. A lot of crucial information such as the salary and previous co-op students’ experiences is missing from these postings. The motivation for this project is to provide a platform where these students can gain accurate and complete information about co-op jobs from previous students prior to applying to those jobs.

Companies are investing in the integration of hardware accelerators to transparently improve the performance of critical, compute-bound code (ex. image processing, log analysis). However, current systems typically tie one software system to one accelerator. This lacks flexibility and provides less speedup than there could be. We designed a tool which allows multiple hardware accelerators to be controlled by a single software system. We have a host side library that distributes work to several FPGA on a local ethernet network.

Oral communication and presentation skills are some of the most important skills for both personal and professional success. However, it is difficult to hone these skills without the help of another person to listen and give feedback. The goal of this project is to create a mobile application that provides helpful insights on presentations. Users can record themselves performing their presentation then the audio and video is processed through machine learning models to provide beneficial feedback such as facial expressions and word choice.