The chosen few: We asked Canada’s business leaders to nominate the brightest minds you’ve never heard of

Frances Donald

Open this photo in gallery:

Macroeconomist at Manulife Asset Management and anti—dismal scientist

It’s no accident Frances Donald sounds fresher and more down-to-earth than other economists who appear on business TV to decode interestrate announcements and government data releases. The 33-year-old head of macroeconomic strategy for Manulife Asset Management says that too many people, especially millennials, are intimidated by economics and finance “because of how these subjects are taught. They’re dry. They’re very quantitative. They’re, frankly, boring.”

So, what is Donald doing to spark interest? Humour helps, and there's plenty of it on her Twitter feed. One recent post includes the question: “Who's your favourite Powell"—Federal Reserve chair Jerome or Toronto Raptors shooting guard Norman? She also tagged news that Rolling Stones guitarist Keith Richards has given up drinking as an early recession indicator.

More seriously, Donald considers herself Exhibit A in proving you don’t have to be a “human calculator” to succeed in finance. Born and raised in Montreal, her mother was a psychology professor and her stepfather sold bonds but also owned an art gallery. Her father was a sales executive for a forklift company. “I thought I’d be a professional violinist,” Donald says. “I practised every day for about 18 years.” Her worst subject at school: math. “I had a tutor every week from Grade 9 until my last year of grad school.”

But reading Freakonomics inspired Donald to major in economics at Queen's University. “You learn political science, geography, psychology, behaviour and history,” she says. “It's a great foundation for a career not just in finance, but in a broad array of industries.”

Effort and chance pushed Donald forward after graduation. She worked at the Bank of Canada as a research assistant when the financial crisis blew through, from May 2008 to July 2009. “It was actually a great time to start,” she says. “So many previous economic relationships and models were breaking down. Essentially, almost everyone was a rookie.”

Donald then earned a master’s degree at New York University and returned to Canada in 2012 to take a job with Montreal-based trading firm Pavilion Global Markets. “This was the first time I had one-on-one contact with investors—with hedge funds and pension funds—and got a sense of how we can use economics to make decisions about finance,” she says. She also wanted to be with her husband, Montreal-born visual artist Eric Clement. In 2014, Donald landed a job with Scotiabank in Toronto as a financial economist, then moved to Manulife in 2016. In September 2017, the couple had a son.

Much of Donald’s job is pure Bay Street. She says she tells portfolio managers “what will happen next and how markets will respond.” That advice is based on rigorous data analysis, but the continuing uncertainty since the financial crisis requires a sociopolitical overlay. “That means constant adjustments and sometimes letting go of methods that have worked for decades,” she says.

But Donald also tries to maintain a long-term perspective on how finances affect people. “I’m a renter like everyone else my age,” she says. Her family lives in the top two floors of a duplex. “My neighbour can hear my son running across the floor, and I can hear her singing.” If more economists were like her, she says, maybe they’d view housing affordability as a social problem, not just “a national financial stability risk.”

Veteran policy economist Armine Yalnizyan is encouraged that Donald has climbed the corporate ladder and still maintained that perspective. She hopes Donald is a harbinger. “There aren’t many women in economics,” Yalnizyan says. “That’s starting to change in classrooms, but it’s unclear whether it will translate into the field.” – John Daly

Leah Lawrence

Open this photo in gallery:

CEO of Sustainable Development Technology Canada and Ottawa’s most efficient bureaucrat

Leah Lawrence wasn't sure how well she'd adapt to life in bureaucratic Ottawa when she made the move east from Alberta. As the new head of Sustainable Development Technology Canada (SDTC), she would be leading the country's largest funder of cleantech startups. One of the things that most surprised the Edmonton-born, Saskatchewan-bred Lawrence was how often she found herself using horse analogies. “This is my first rodeo. You can lead a horse to water—I never said any of those things when I worked in Calgary,” she says, “but weirdly, they started coming out of my mouth” after she arrived in the capital.

Here's another horse analogy: Since taking the reins at SDTC four years ago, the 49-year-old engineer has transformed a moribund government agency into a vibrant cleantech player, despite being a rookie bureaucrat.

Before Lawrence came along, SDTC operated much like many government innovation programs, funding projects that promised jobs and reflected government policy, rather than focusing on whether recipients addressed a real market need. Now, she says, SDTC asks questions like: “Does the customer want it? What's your intellectual property strategy, and what does the market look like?” Creating green technology is the goal, but “if it doesn't get deployed, it doesn't matter, which means there has to be a business model to support that big idea.”

In addition to becoming more market-focused, SDTC—which has provided $1.5 billion to about 300 companies over its 17 years, and plans to disburse almost $100 million this year—has earned high marks from the Treasury Board Secretariat for being client-focused and cost-effective. The agency's operating costs are 11% of allocated funds, compared to an average of 29% for federal programs. Applicants used to wait two years for funds—that's now down to six months, and Lawrence wants to shrink it to four.

Meanwhile, she has transformed SDTC into a champion for young companies, educating them on the importance of intellectual property, data and scale-up strategies, and helping them navigate a globally competitive sector. SDTC under Lawrence is “an advocate not just of the companies they invest in, but also trying to have a broader impact in Canada,” says Hamid Arabzadeh, CEO of Ranovus Inc., an SDTC-funded Ottawa startup that's looking at ways to make data centres more energy-efficient.

Modest to a fault, Lawrence doesn't like to dwell on her professional accomplishments (she chaired Calgary's Chamber of Commerce and ran a cleantech advisory firm for 12 years). Instead, she prefers to highlight those who've influenced her, including emissions trading pioneers Carlton Bartels and Adam Smith, who “didn't live to change the world” after perishing in the attacks of Sept. 11, 2001.

And while adapting to Ottawa, she has also brought a bit of the West with her, encouraging staff to wear jeans to work on Fridays and throwing her annual Calgary Stampede party in the capital. As Lawrence puts it: “I find they don’t really have an opportunity to come out and just have pancakes and enjoy each other’s company, with no other expectations than having pancakes and enjoying each other’s company.” – Sean Silcoff

David Duvenaud

Open this photo in gallery:

Founding faculty member at the Vector Institute for AI and a pioneer in teaching machines to learn

Maybe it was intuition or maybe just indecision. But at 22, David Duvenaud knew it was time to switch tracks. He’d finished his computer science degree at the University of Manitoba, but failed to win a co-op placement or get into grad school.

So while his friends headed off for higher learning, he stayed home in suburban Winnipeg, where he killed time building a giant tower out of Lego and ploughing through War and Peace. He considered planting trees, but having spent his youth doing chores on his family's farm, he knew he loved fresh air but not the backbreaking physical labour. Instead, he joined the Army Reserve.

Fast forward 14 years, and Duvenaud finds himself looking back not on a military career but on an academic track that has led to one of the most vibrant research hubs in Canada. A founding faculty member of the Vector Institute for Artificial Intelligence in Toronto, he has gained international attention for his work in an area known as approximate inference. “You could say we're trying to automate the process of building intuition,” says Duvenaud, now 35.

In December, he won a “best paper” award at an annual showcase for the latest in machine learning. The project, which began as an effort to predict patient health, turned into a new way to equip computer systems to deal with information that comes at unspecified intervals, rather than on a set schedule. “David combines very strong mathematical and technical knowledge with a love of working on code, so he knows his own work and the field inside and out,” says Richard Zemel, the Vector Institute’s research director.

Things could have unfolded differently. In 2006, Duvenaud was working toward a tour in Afghanistan when friends from high school invited him to join their startup. Duvenaud threw himself into the technical side of the company they dubbed Invenia, which included optimizing the performance of electric grids. It was his first taste of deep learning, a tool developed in part by the University of Toronto’s Geoffrey Hinton (a big fan of Duvenaud).

To learn more, Duvenaud finally obtained a master's from the University of British Columbia and then attended an immersive AI summer program in Toronto. That's where he met Christian Steinruecken, now a researcher in machine learning at Cambridge University. “David easily stood out as one of the smartest attendees,” says Steinruecken, who suggested he do a PhD in Britain.

Duvenaud heeded the advice and attended Cambridge, where he studied under Carl Rasmussen, one of Hinton’s former students. (He also met his future wife, Melissa Trujillo, a sociologist.) Following a stint at Harvard, Duvenaud was drawn back to Toronto and the Vector Institute, where he has been working on how computers can learn to anticipate effects without knowing their causes. The feat is akin to the way a doctor decides what’s making a patient sick, based on a collection of symptoms.

“The most satisfying part of research in machine learning is that it helps us understand how and when knowledge is possible, for humans or machines,” says Duvenaud. “Trying to get machines to learn forces us to carefully and explicitly think about learning…and the best part is that there’s no way to bullshit through the problem.” – Ivan Semeniuk

Humera Malik

Open this photo in gallery:

CEO of Canvass Analytics and saviour of industrial dollars

For Humera Malik, it was a pretty typical week: First, she flew to Davos to pitch her company’s predictive analytics software to industrial giants at the World Economic Forum, where the global elite gather to tackle the world’s problems. Then it was off to San Francisco for back-to-back-to-back meetings with investors, who have put more than $5 million (U.S.) into Malik’s Toronto-based startup, Canvass Analytics.

Malik has a long history of making old things work in new ways, from building early WiFi networks to, more recently, using artificial intelligence to make better decisions on an industrial scale—decisions like how much power your factory is consuming or how you monitor product quality or even when you need to service your equipment. Canvass, which Malik founded in 2016, harnesses machine learning to help metals, auto, agriculture and energy plants adapt to a rapidly innovating world. These industries, she says, tend to be “data-rich but information-poor” and often need a nudge to use data to their advantage.

The 44-year-old entrepreneur, originally from Pakistan, studied computer science at Georgia Tech and worked for U.S. telecoms before joining Bell here in Canada. Her various roles took her all over the globe. “I was able to immerse myself in different cultures,” she says, “which gave me insight into other management approaches, and even how they perceive and adopt technology.” As a consultant earlier this decade, she began working with industrialscale connected devices and sensors, and soon realized how much data factories and plants produced but left uninterpreted. Canvass's AI makes sense of it all.

For one metal processor, Canvass uses temperature and raw-material composition data to predict the quality of final products and reduce waste. Another client was running its natural gas turbines at full capacity 24-7 no matter its electricity demands; Canvass's software was able to predict energy demand so the company could cut down on fuel consumption and wear and tear. And in November, Malik's company signed a deal with Singapore agri-giant Olam International to make its supply chain more efficient.

Malik has attracted many of her high-profile investors from both the industries Canvass serves and the technology it embraces. The company’s latest round of financing, in August, was led by Gradient Ventures, the AI-focused fund owned by Alphabet Inc., with help from Bedrock Industries, a major shareholder of Stelco. Among the first to buy into Canvass’s vision was Real Ventures partner Janet Bannister, who met Malik in 2017. By that point, the young company was testing its software in Canada and Europe. “What impressed me was her determination, her hustle and how much she had done,” says Bannister. Canvass, she adds, “could be a billion-dollar company.”

For his part, MaRS Discovery District CEO Yung Wu thinks Canvass could have “trilliondollar impacts.” For many manufacturers, “every efficiency you can drive is life or death,” and he sees Malik as a key leader in that transition. As Malik notes, food and energy demands are expected to grow, while many legacy industries’ institutional knowledge is drying up as boomers retire. AI can help the next generation build upon the last. “It’s not a matter of if industry will,” she says. “It’s a matter of when.” – Josh O’Kane