With $200 million of capital available, Business Development Bank of Canada Capital’s Deep Tech Venture Fund is the country’s largest fund dedicated to deep tech.

The fund focuses on early-stage seed and Series A equity investments, with a 12-year life and the possibility of a four-year extension — longer than the 10-year cycle of a standard venture fund. Deep tech start-ups benefit from “patient capital” that allows more time for research and development, says Thomas Park, the lead partner on the fund.

Park says the fund is looking for deep tech start-ups in areas like quantum information sciences, photonics, electronics and foundational artificial intelligence. Deep tech refers to companies built around substantial technological or scientific advances that are for other companies to reproduce. So far, the fund has invested in two quantum technology companies, Xanadu and North Quantique.

Research Money spoke to Park this week about the fund and the future of deep tech investing in Canada. Before joining BDC in 2016, Park was a program officer for The Gates Foundation, where he worked on overhauling Senegal’s contraceptive supply chain. He has also worked as a senior engagement manager for McKinsey and in positions at the United Nations Assistance to the Khmer Rouge Trials, the OSCE Mission in Kosovo and the International Criminal Court.

You left your position at BDC Capital’s VP of operations and strategy to become head of this Deep Tech Venture Fund. What was the problem you saw then that you wanted to solve? 

I was quite involved in a lot of public policy discussions on innovation, especially the venture capital part. There was consistent noise that we were hearing that there was a gap in transforming Canada's R&D efforts and efforts to commercialize it. We've heard this for a while. But it became even louder because there was concerns about certain key technologies, such as quantum tech, foundational artificial intelligence, photonics, advanced materials, electronics. Look what's happening to AI. AI, pre-2010, was considered deep tech. And then it exploded, but there wasn't much of an ecosystem to support the commercializing of Canadian R&D efforts.

That's what we're trying to test here. Can we actually commercialize a lot of this difficult tech that requires patient capital.

I’ve seen you refer to this type of investing as “patient capital.” Can you explain that?

Good question, because every VC fund will say they are patient capital [laughs]. The problem was that the typical VC life cycle of 10 years isn't long enough for a lot of deep tech startups. For the first five years the fund will invest and in the remaining five years they'll look to exit. You can't build a quantum computer, a new chip or even some of the new AI tools that are coming out within a five year period. And so we're a twelve-year fund with a possible extension of four. We're really designed to be this much longer-term player in terms of how we invest. And that's what's needed for a lot of these deep tech companies. They just take much longer to commercialize.

Tell me about deep tech. How do you define a deep tech company? How do you judge if something falls in that category?

Deep tech is where there is significant technical risk. We are looking for fundamental breakthroughs in science.

Is there a strong affiliation with a very well-respected academic? Usually they play an advisory role or often the part of chief science officer. That's one [element]. And two, that there's some breakthrough in the science. If you take AI for example, we look for innovations in the algorithm and not in the data set. That's key there.

What's not deep tech?

The overwhelming number of start-ups that we have which, by seed, already have a software product and they just need money to start selling this product. Think of apps trying to connect doctors with patients or things like that.

What are examples of the transformative technologies you're looking for?

We have made two investments: one in Xanadu and one in North Quantique. We're really excited about explainable AI. If you think about it, we were deploying all of these AI models for the last several years — neural nets, etc. Very exciting, but we didn't actually understand why they were making these decisions. You can't actually deploy these models, especially in a bank...unless you can explain how it came up with this decision and how it deals with all the different potential biases.

We want an algorithm to have bias. This is a cat, not a cat. We just don't want it to have the wrong biases. That's what explainable AI does. It goes back into the neural network and says, this is how it came up with a decision. Otherwise it's too complicated. It'll take you months to explain a model.

Part of what you're doing is trying to predict what technologies will make sense in the 2030s. Isn't there some risk there? How do you find technologies that will be valuable so far in the future?

We look for a few things. One, is there a market demand? With quantum computers, we believe there is a market. It could be far off, but if they actually build it, there will be a market. On the shorter term, in explainable AI, there is a demand right now and we just see it increasing.

The second thing is, can the team over the next 10 years come up with a product/market fit? We can test it out with government procurement, for example. We help them navigate Innovation Solutions Canada and all of these different programs and see if they can actually solve problems.

The third thing is the team. That's why the quality of the management, especially about the affiliation with research institutes, is key. Because they'll need to maintain the edge. As soon as the world finds out what they're doing actually has promise, they will create their own start-ups. Can they actually maintain their edge technologically over time?

Where do deep tech start-ups usually fail?

One of the biggest problems is where they say 'I've got a technology,' but they haven't found a market for it and can't articulate what the market is. What is the pain point? We see that a lot with edge AI chip companies. They'll say I've got a great chip and it'll be on devices rather than on a cloud, and it'll be on your iPhone or on your printer or wherever. But the problem is, is there demand for that chip? What's the problem it's trying to solve?

The second is management. What's unique about deep tech companies is that you usually have a researcher that is affiliated. How much time are they actually spending on the company? Because someone is a rock-star academic doesn't necessarily mean they are going to be an amazing co-founder. What role are they actually playing? The incentives that an academic has are very different from what a founder of a start-up has. And they often won't leave their university position — fair enough — so what role are you going to be playing in the start-up to set it up for success, as opposed to the start-up being one in a portfolio of other initiatives you're doing as part of your academic career?

Third is IP. Who actually owns it? You start up the company and then realize that the university actually still owns all of the IP. So what are we investing in? Those are the three big buckets: IP, management and the market.

Canada is relatively weak on commercialization compared to ideas and research. A lot of people have weighed in on the systemic reasons for that, from better protecting IP to a lack of focus from our government. From your perspective, where are the low-hanging fruit where we can take action on right now?

A low-hanging fruit is IP policies at a lot of universities. A good example is the University of Waterloo. They have a much more flexible IP policy, and it's not surprising that the university became such a strong spot for entrepreneurship. I don't blame the other universities, because what's interesting is that other universities have medical schools, large faculties in medicine, and I can see the hesitation about more flexibility with IP. A lot of them have been burnt by deals in the past, they've lost out on medicines and medical devices, and so I understand.

Two is a mindset shift. We can protect our IP as much as we can but unless we do something with it, it's not very valuable. You can sit on a bunch of patents but unless you build a management team to actually do something with it, it isn't as valuable as you think it is... we do need a more comprehensive, holistic and at least clear strategy in IP, but we need this other component where there is a group of people who can do something exceptional with that IP.

You have an interesting history of working on global health and development projects, from Senegal’s contraceptive supply chain to many years ago clerking with the United Nations at the Khmer Rouge Tribunal. What’s the through-line there for you?

One, it's about giving back. It's service. I feel very privileged, so I have an opportunity to leverage my education and skills to help others as much as I can. I would say the second thing is my outlook on how do we solve the wicked problems of the world? Before it was about human rights legislation and building a global human rights system, and then global health and realizing how that's done. Canada punches above its weight but we can do a lot more, especially about innovation. What's nice about Canadian innovation is that it is about Canadian values. If you look at Element AI and a lot of the AI start-ups here, they created a Montreal AI Ethics Institute and a Montreal protocol. That's a very Canadian thing to do.

The problem is that so much of the tech we have is publicly researched but the gains are privately held. If you look at the internet and mobile technology, those are publicly-funded research technologies but all of the gains are captured [privately]. For BDC, if start-ups become a unicorn and they exit, that's great for the start-up founders, it's great for investors and great for citizens because the gains are recycled back into the system. Canadians can profit from the upside of the gains from investment in public research. I'm really quite a fan of Mariana Mazzucato's work [an influential economics professor at the University College London] — these development banks allow citizens to gain from the upside of the technology that citizens have invested in.

This interview has been edited for length and clarity.

R$