RBC and the Borealis AI
lab are continuing their push into one of Canada’s strongest tech sectors.
Two new top AI talents are joining Borealis AI research centres: professor Pascal Poupart from the University of Waterloo and professor Marcus Brubaker from York University. Both will join as senior researchers.
“It’s exciting to now have opportunities for researchers like these who have worked within Canadian academic institutions continuing to pursue research within the Canadian industry—not just on Canadian soil, but for Canadian companies too,” Foteini Agrafioti told Techvibes. Agrafioti is the chief science officer and head of Borealis AI
“That wasn’t the case a few years ago, but now we’re at a tipping point. The fact that Marcus and Pascal want to stay here and join the Canadian effort, it means we’re doing something right,” added Agrafioti.
Poupart will split his time between teaching at the University of Waterloo and research at Borealis AI. His team will be based out of Waterloo, and his background in the development of algorithms for reasoning under uncertainty and machine learning with application to assistive technologies, natural language processing and telecommunication networks.
AI would not be the industry it is in Canada today without Poupart. His research and work with companies like Google, Huawei, Intel and others focused on real-world applications and the expansions of AI into financial services.
Brubaker is an expert in computer vision and probabilistic models. He works on problems where uncertainty plays a major role and where answers and inputs aren’t always clear. Brubaker’s goals often include finding out how to work and make predictions within these frames of uncertainty.
It was his research in the cryo-EM field that went on to heavily influence the winners of the 2017 Nobel Prize in Chemistry.
Part of the fundamental research Brubaker is undertaking will be in the realm of removing bias from AI and machine learning models. In addition, he will work on explainability, a topic that can lead to increased transparency in how decisions are made in machine learning.
“We’re looking at how to take a lot of the great advances in AI and machine learning over the years and endow them with a sense of that uncertainty and how to manage and make predictions about that uncertainty,” explains Brubaker. “Then we can develop algorithms that not only provide good answers but also understand when they don’t know the answer.”
In this sense, the financial world can utilize AI in several unique ways, many of which do not specifically involve cash flow but tie into how more overarching and complex problems are solved. The goal is to help remove boundaries for customers to access the product while at the same time removing possible regulatory hurdles down the line. Brubaker and Poupart are extremely well-qualified to take on these challenges.
“Marketplaces are extremely complex,” says Agrafioti. “For us to be able to protect our clients and create products that are meaningful in such complex and ambiguous environments, these types of AI are extremely important. For Brubaker and Poupart, it was a great fit both from a fundamental and academic research perspective, but also for application of their research in the financial services industry.”
“If we as researchers can’t give good answers to these questions like bias in algorithms or how decisions are made, we’ll start facing stiffer regulations from governments that are concerned about their impact,” adds Brubaker.
Borealis AI is an RBC Institute for Research dedicated to achieving state-of-the-art machine learning advances. They are heavily involved in collaboration between educational institutions and industry and highly promote their work with schools such as York University and the University of Waterloo. Keeping lanes between schools and business means they can easily tap into the growing AI networks in Canada and pique the interest of younger generations looking to make an impact in the space.
Between the Toronto and Edmonton Borealis AI locations, there are roughly 45 total researchers, and the new Montreal location will open in the spring. When that happens, expect some more big hires from RBC and Borealis AI.