FIVE TO WATCH

Michael Katchen, founder and CEO, Wealthsimple Inc.

Canada’s entire existence has been dominated by monopolies and oligopolies, which has provided a bit of a safety net, but has also stifled competition and limited innovation, perhaps nowhere more so than in the financial services sector. But that didn’t stop Michael Katchen from launching an automated investment service online — now more popularly known as a robo-adviser — that didn’t require a lot of hassle to open. That was five years ago and, of course, there are many more robo-advisers in the market today, but Toronto-based Wealthsimple keeps pushing the envelope to offer cheaper, sometimes even free and arguably better banking-type services such as commission-free trading and high interest savings accounts, forcing the Big Banks and financial services players to adapt.

“The company got started out of an organic need. I wanted to help some friends invest and there wasn’t a good option out there so I decided to solve it for them,” Katchen says. “It is so important to be smart about money and to be thoughtful about the fees you pay on the services you get to manage your money, to enable you to live the life that you want and to provide the life you want for your family. The financial service industry has done a disservice to Canadians by not pushing the envelope.”

Pushing the envelope could almost be Katchen’s mantra. For instance, Wealthsimple in 2016 won an award for having the world’s best financial services company website. The next day, the website was completely torn down and redesigned from scratch. “Our belief is that innovation and iteration go hand in hand and they’re never done,” he says. “It’s an ongoing process.” Wealthsimple has also pushed abroad, opening operations in the United States and the United Kingdom, which have more robust players and competition.

It must be daunting, especially initially, to take on big established players even in a country such as Canada that is begging for better service, but Katchen doesn’t seem to feel it. He doesn’t think about being an industry disruptor, but he does think about raising the industry’s bar. Either way, Wealthsimple’s message seems to be resonating with consumers. Banks

have responded by offering higher interest rates, and trading commissions are dropping, sometimes to zero. “One of the things we’re excited about is that we’re a five-year-old company so we’re still brand new, but we already are moving the industry forward,” he says. “Even though we’re competing against some of the largest financial institutions in the world and the Canadian banks here, we continue to keep moving fast.”

Speed is the key since Wealthsimple has but one patent to its name. The company, similar to Shopify Inc., is evolving without worrying too much about whether someone will mimic its technologies and service offerings, especially since Katchen believes patents are just a little passé given how fast markets are evolving. Where once a patent, generated after years and millions of dollars’ worth of research, could offer protection for 10 or 20 years, today it might be out of date before the application is even dry. “Our belief is that it’s the speed at which you iterate and innovate that is the way you keep a competitive advantage today.”

All of which is why Wealthsimple keeps adding services. It just launched a tax service, SimpleTax, which it expects will help one million filers this year, and Katchen names responsible credit, such as houses, mortgages and student debt that help people augment their human or wealth capital, as targets. And then there’s insurance. “We’re really in the square of saving and investing today, which is really the first rung of building good financial behaviours and financial freedom over the long term,” he says. “We’ve got an enormously exciting road map in front of us.”

Kurtis McBride, co-founder and CEO, Miovision Technologies Inc.

Like it or not, and many don’t, especially in Alberta, Canada is part of the Paris Accord that pledges to reduce greenhouse-gas emissions by 30% below 2005 levels by 2030, and it’s here to stay — at least until a government change anyway. Reducing our dependence on fossil fuels is tough, to be sure, but there is a Canadian-made solution that could take care of a third of that commitment and it wouldn’t take much effort at all. Just solve something we want solved anyway: traffic congestion. Rather than spend millions, if not billions, on road infrastructure, Miovision’s software can simply make Canada’s tens of thousands of traffic intersections operate more efficiently, cutting the time we spend idling and polluting the atmosphere. “The opportunity to have meaningful impact on people’s quality of life, the livability of our cities, the environment is real,” says Kurtis McBride, co-founder and CEO of the Kitchener, Ont.-based company.

Reducing congestion is such a simple, yet popular idea that it’s a wonder no one seems to have thought of it before, but McBride says he’s always questioned why things are the way they are, and traffic was one that came to him as a co-op student at university. “Most people assume traffic has always been bad, it will always be bad and there’s really nothing they can do about it, but I don’t think that’s true at all,” he says. “When we started the company, it was really on this idea that if we just applied ideas, concepts and business models and technologies that had been proven out in other markets for completely different problems to traffic problems, then traffic doesn’t need to be this bad.”

Put simply, Miovision takes 21st century technologies such as the cloud, connectivity and edge computing and applies them to collecting data from the traffic boxes that control the lights, most of which haven’t been thought about too much since the 1970s. “Once the bridge between those two is built, then everything else we do is in software,” McBride says. The data that is collected by the Miovision cameras is owned by the municipalities that operate the traffic lights, and can help them figure out how best to operate those lights to maximize efficiencies.

Two fundamental problems occur in a traffic network. One is called a split failure, or over-servicing one direction and under-servicing the other, causing unnecessary backups. The other is a coordination failure, which is where drivers hit every red (or every green) in one direction — nice when you hit green, less so with the red. Data can determine where those failures are occurring and then Miovision, using its tools or third-party ones, can implement a plan, whether that’s based on time of day, day of the week or even seasonally, sensors or adaptive control, which McBride says is almost a real-time network-level response.

“Think of us like a smartphone of the intersection with an app store, so we have lots of native apps and partner apps,” he says. “We give the municipalities access to the raw information that they can access through APIs. We focus on congestion mitigation, improving safety and improving operations, so everything from if the power is out at a signal and creating an unsafe situation all the way up to things like fixing green time.”

The one impediment, it turns out, isn’t the municipalities, which McBride says get a bad rap for not pushing innovation, but the procurement system that lends itself to vendors selling closed systems, so there’s no need to ever get better. Municipalities, of course, could drive open standards, “which would open the market up to innovation, but vendors could do a lot more to drive innovation as well,” he says. Open systems would have the side benefit of reducing some of the privacy fears that having private-sector cameras on every intersection will no doubt generate. McBride says it’s up to the cities to determine what to do with the data, though he recognizes that there’s a branding issue since it’s Miovision’s name on the cameras. To that end, Miovision, along with Esri Canada, Geotab, MappedIn and others, created an organization called the Open City Network that addresses the need for openness when developing smart-city infrastructure. “We don’t collect any PII (personally identifiable information). We go out of our way to make sure we don’t capture any PII,” McBride says. “Our interest is in making traffic flow better and safer.”

Graham Taylor, associate professor, School of Engineering, University of Guelph

Robots get all the attention when it comes to artificial intelligence and perhaps that’s not out of line, given the fear they will eliminate millions of jobs, particularly blue-collar ones. But the power of AI is a lot more than just riveting one part to another, or answering simple queries from confused customers. Indeed, there’s growing interest in Canadian AI circles, Graham Taylor says, to democratize the technology, for example, by automating database queries. “The holy grail would be to have a spreadsheet and you’re familiar with working in Excel, but you’re not a programmer and there are some missing data, and you could just say to the algorithm, hey, fill in the missing data,” says the University of Guelph professor, startup consultant and all-round AI expert. “It would then figure out where the data is, crawl the web, write queries in databases, and do whatever it needs to do to fill in the data.”

That doesn’t sound nearly as threatening as robot overlords, but democratizing AI is something Taylor is quite interested in. Traditionally, AI, even to do something seemingly simple as generate or recognize faces, requires huge computing resources that only a handful of companies have at their disposal. For example, BigGAN (Big Generative Adversarial Network) needs 512 TPUs (tensor processing units), which is specialized equipment that only Google has. Although Taylor says Big Tech has been fairly good about showing what they’re doing, even sharing their codes, small companies and academics have been left out due to the cost. “What we’re focused on specifically is not reducing the number of examples that you have, but compressing the examples that you use to train the computer,” he says. “We’ve been able to run it on a single GPU (graphics processing unit) that costs about $1,000.”

Perhaps more fun is the work Taylor and four undergrad students at Guelph are doing to build a generative model that creates new Lego designs based on the bricks at your or your children’s disposal. He’s also interested in using machine learning systems and software to do customization at scale. Many engineers use the 80-20 rule when creating products: that is, design something that will work for 80% of the population. But Taylor thinks it will be possible to use AI to reach the other 20% with mass-produced customized products, as opposed to trying to retrofit or build something entirely different. Again, the idea of AI being a democratization agent at work.

Of course, there will always be suspicion of AI’s motives, or at least the motives behind the people and companies that use AI. That, too, forms a big part of Taylor’s work. “There is a lot of hype positively, a lot of hype negatively and a lot of misinformation out there,” he says. In 2018, the University of Guelph started the Centre for Advancing Responsible and Ethical AI (CARE-AI) to help address some of that fear and anxiety. The centre hosts public information seminars and generally tries to insure the ethics surrounding AI don’t get lost in all the excitement. Taylor is also academic director of Next AI, a four-year-old founder

There is a lot of hype positively, a lot of hype negatively and a lot of misinformation out there

development program that brings in experts from around the world to get entrepreneurs thinking about the ethical issues. “In the first and second year of the program, we weren’t getting a lot of questions about ethics and responsibilities from the teams, but in the past couple of years it has really spiked up,” he says. “It tends to one of the areas that the teams want to know the most about.”

Though Taylor co-founded Kindred Inc., which is trying to make robot workers that better emulate what humans do, he discovered balancing academia and a full-time business was a bit overwhelming. “Running a company is more than a 100% job,” he says. “University is more than a 100% job.” That’s not to suggest you won’t find Taylor in the trenches. He did a sabbatical at Google Brain last year, remains an academic adviser to Kindred and collaborates with startups such as Toronto-based Swift Medical Inc., which has created a mobile app for wound care (bed sores, diabetic ulcers and the like) in nursing homes. “I really enjoyed all these experiences, but when I do them, I really feel my heart is in academics. These experiences have been amazing, but the best place for me is in university.”

Joshua Wong, founder and CEO, Opus One Solutions Energy Corp.

If you’re going to have a big hairy audacious goal, it may as well be the biggest, hairiest, most audacious goal, and Joshua Wong, CEO and founder of Toronto-based Opus One Solutions has one: to be the Amazon.com of energy. Opus One can already collect data from power grids to help customers digitize the information they collect, hopefully allowing them to make better operational decisions, thus reducing the amount of wasted power and decarbonize. But the ultimate goal is to decentralize the grid or empower individual homeowners and business owners to participate in both buying and selling electricity.

“With Amazon, you can trade consumer goods, but they still need the highways, shipping and logistics to ship the goods. We still need the poles and the wires, but we can transact energy like e-commerce on top of that traditional infrastructure,” Wong says. “That’s the final solution to all the debates, in Ontario especially, about fees and tariffs, and the U.S. and others that have net metering subsidies for clean energy. Energy should not be subsidized, it doesn’t need to be subsidized, you just need to use the data analytics to drive the right pricing for energy.”

Data is the key, Wong says, because the infrastructure is in some case 100-plus years old, and while the resulting grid may be the most complex invention of the 20th century, it needs help to achieve the next phase. And there’s nothing cheaper on the grid than data, especially when compared to the cost of running equipment down and not adding new energy sources and users, or replacing it with current, but expensive newer equipment. “Data allows you to not only be the most efficient as to how you manage the infrastructure, and hopefully reduce rates and improve reliability, but it also allows you to, in the most informed fashion, choose between traditional investments, which is more poles and copper wires and oil, or the new stuff, being solar, storage, electric-vehicle charging, automation of components. It’s the analytics that inform how we should evolve the infrastructure.”

For example, it’s possible to create microgrids, self-sufficient silos or islands that can be as small as one house, but that isn’t very efficient. Inter-connecting these microgrids into a bigger whole, however, can improve resiliency, a fancy way of saying avoid outages. Opus One recently set up such a grid that connected Tesla batteries to store wind power in Nova Scotia. It was able to support the area for 19 hours when Hurricane Dorian hit in 2019 and shut the power off everywhere else. “That’s the power of the sharing economy of power,” Wong says.

Microgrids allow utilities to incorporate new forms of power such as solar and wind, as well as storage batteries. But data is still needed to figure out where best to put such equipment and how it should be controlled.

Opus One is also working with hydro utilities in Ontario, as well as some of the largest utilities in the United States and around the world, and breaking into the Asia Pacific market. “The market is ready and ripe. The data sets do succeed,” Wong says. “The biggest thing is further education, some regulatory support is always needed, but it’s also good that, fundamentally, the market is moving to a different stage: how best should I use these new tools?” He has a few ideas.

Michael Tamblyn, CEO and president, Rakuten Kobo Inc.

Canada’s history of creating cutting-edge products is rich. Making them successful? Not so much. There’s the light bulb, Avro Arrow and BlackBerry to name a few that were punished by market forces. Then there’s the e-reader. Toronto-based Kobo certainly didn’t invent it — the idea has been around since a 1930 manifesto called The Readies written by American writer Robert Brown — but it has popularized the device even though it’s up against direct competition from Amazon. com Inc.’s Kindle as well as indirectly from tablets and smartphones by the likes of Apple Inc. and Samsung. Kobo now operates in 24 countries, making it one of the largest e-book retailers and device manufacturers.

That Kobo is still kicking 11 years after it was founded counts as CEO Michael Tamblyn’s big surprise. “We started with the idea that there would probably be three or four companies that would dominate e-books and e-reading globally and it was certainly our ambition to be one of them,” he says. “Through some choices, good bets and good luck, we find ourselves in a place where it’s those three big ecosystem players that are big everywhere — Amazon, Apple and Google — and then there’s us as the one big pure-play player in e-books and e-reading.”

Marketing is one of the keys to innovation that is often overlooked, and Kobo would not be a success without it. In particular, Tamblyn says Kobo works with existing retailers, rather than try to disrupt their models. It’s a tactic that makes sense when you recall Kobo (an anagram of book) is an offshoot of Indigo Books & Music Inc., though it is now a subsidiary of the Japanese e-commerce conglomerate Rakuten. In the 2000s, Indigo, like other book retailers, was grappling with how best to accommodate the growing desire for digital content, something the music and video industries struggled mightily with as well. But, after all, if you want to find someone who loves books, go to a bookstore.

“It’s been interesting to watch the book industry navigate this change in a very different way than in music or video, where it’s been much more of an all-or-nothing transition,” Tamblyn says. “We see publishers stepping up and making certain kinds of books more beautiful so that they earn places in people’s homes and, at the same time, embracing digital distribution because somebody who is an avid reader and churning through two science fiction books a week is someone you want to keep reading digitally as fast as possible.”

Aside from Indigo, some of the other retail partnerships Kobo has include Waterstones, the largest book retailer in the United Kingdom, Libreria Porrúa and Gandhi in Mexico, and Hugendubel, Thalia and Weltbild in Germany. They get access to a catalogue of more than five million e-books, newspapers and magazines. In Canada, about one in five books read for personal use is currently read in digital form, which Tamblyn says is a fairly significant penetration rate when you consider how durable print is as a format. Typical users are as likely to be 65 years old as they are 18, he says, and Kobo says it already has 12 million users in 190 countries. At some point, “we probably won’t be pressing ink into dead trees to do our reading,” he says. “Everything after that is a just question of timing.”