How an undergraduate student team supported by RBC Borealis used Machine Learning to help solve a global communication problem.

The Let’s SOLVE it mentorship program is now entering its second year. A joint program by RBC Borealis and CIFAR, the program is aimed at helping students gain valuable machine learning (ML) skills, so they can solve real problems in their communities that they care deeply about. Let’s SOLVE it program is open to all undergraduate students at all Canadian universities and is conducted virtually – teams don’t need to be in the same location or at the same school in order to participate. The students don’t need to be enrolled in a Computer Science program – while team members would benefit from having some basic programming knowledge, specific experience using AI or ML isn’t necessary.

Student teams get to pick the problems in their communities they would like to solve.

More than 100 million people worldwide are affected by hearing impairments or speech difficulties.  And American Sign Language (ASL) is one of the most used sign languages in the world. Yet some estimates suggest there are fewer than 500,000 users. The opportunities to freely communicate are often limited.
 
It’s a Friday afternoon, and five Carleton University students are talking to a Simon Lemay, Machine Learning Engineering Lead, RBC Borealis about their big idea to solve this problem. They want to build a Duolingo-like app that can make learning ASL fun and intuitive.

🤝 Meet the student team: How Hungry Hippos got from spotting a problem worth solving to making an impact.

The five students are participating in a RBC Borealis program called Let’s SOLVE it. Throughout May and June 2022, the students – who organized themselves into a team called Hungry Hippos — taught themselves some foundational and practical aspects machine learning. With no prior experience in ML, they watched videos and took online courses to learn more, and worked closely with Simon Lemay, their RBC Borealis mentor, who helped guide their learning and put the new ML skills to work. The team ended up building a demo of their ASL app – getting many steps closer from identifying a problem worth solving to helping make a dent in a difficult and global communication problem.
 
“We came into this program wanting to build a solution to a real-world problem but had no experience with ML/AI. Through the Let’s SOLVE it program, RBC Borealis gave us the opportunity to explore machine learning in our own time while also providing great resources to help us along the way; like computing power to help train models faster,” the team noted in a recent blog post. “The Let’s SOLVE it program was incredible, and we all learned a great deal about machine learning.”
 
The Hungry Hippos team consisted of five members: Sarah Chow, Shuvaethy Neill, Harsimran Kanwar, Hussein Elmokdad, and Guy Morgenshtern, all third year Software Engineering students at Carleton University.
 
They presented their working prototype – called Sign Me Up – to leaders at RBC Borealis, CIFAR and RBC this past summer.

Watch the below video to see an overview of their LSi prototype.

Building the future of ML ecosystem through mentorship.

“The Hungry Hippos’ Sign Me Up project is an excellent example of how organizations can help students learn key ML and AI skills while also enabling them to make a positive social impact,” noted Dr. Eirene Seiradaki, Director of Research Partnerships at RBC Borealis. “At RBC Borealis, we want to expose the next generation of ML and AI leaders to ideas, capabilities and resources that can help them advance their development while also serving the needs of society.”
 
“I always thought ML was something I couldn’t do because it involved a lot of math,” noted one member of the team. “The Let’s SOLVE it program showed me how accessible ML can be.” As another member added, “My experience at Let’s SOLVE it really opened my eyes to the different ways you can apply Machine Learning. The program gave me the confidence to really explore ML as a future career path.”  
 
Let’s SOLVE it is part of a suite of CIFAR NextGen AI Training Programs aimed at diversifying and developing the Canadian ML and AI ecosystem through resources, mentorship and practical training.