Vantage points: Cohere’s Ivan Zhang on making Canada the place to scale

Cohere’s Ivan Zhang believes Canadian tech is great at creating winners, but terrible at keeping them here.

The company he co-founded is a case in point: Cohere, which builds enterprise-focused AI models, reportedly scaled to $22 million in annualized revenue in March and has pulled in nearly $445 million from investors to date.

Yet, as Cohere gears up to reportedly raise another $500 million at a $5 billion valuation, Zhang believes there are gaps in the Canadian ecosystem that have compelled the company to seek talent and growth abroad.

“Being a winner in our class of products means that we can compete with American salaries and keep talent here.”

Like many scaling AI companies, Cohere needs an abundance of capital, talent, and computational resources, all of which are tough to find in Canada. While Zhang remains bullish on what Cohere can achieve as a Canadian company, he doesn’t believe that it’s a country that takes its own tech companies seriously enough.

Fresh off a redeye flight for BetaKit’s Town Hall last week, Zhang shared his thoughts on the country’s talent retention challenges, its little brother syndrome, and why it’s “very hard to stay in Canada if you want to be the best in the world at something.”

The following Q&A contains Zhang’s responses from the vantage points panel at the BetaKit Town Hall and a separate interview. Responses have been edited for length and clarity.

What challenges and opportunities are you seeing at this moment in time in Canadian tech?

I’ll start with the opportunities. Canadian tech has always been early. They’ve always been really good at picking out winners before they become world renowned, and that’s been true of AI companies especially. At some point, most of the heads of AI at major labs were just Toronto students under Geoff Hinton, studying at U of T. I think it’s really good that we have a really good top-of-funnel pipeline.

Where we’re lacking is actually retaining that talent here, and it’s due to multiple factors. It’s hard for a junior engineer to get exceptional mentorship, for example, from Toronto firms. It’s hard for entrepreneurs to grow their business, meet more customers, or raise more funding, just in Canada. They’re forced to expand like we have. We expanded offices to London for both talent and customers.

At @cohere London office for Cohere's Build Day! pic.twitter.com/45P5wVjROz
— Amandine Flachs (@AmandineFlachs) May 8, 2024

We have to work backwards on how we make Canada a place where the best in the world can spend their entire career. Right now, that’s not true. The best in the world spawn here, but very quickly, they hit the ceiling of whatever craft they’re working on and they have to go elsewhere.

I think everyone in the ecosystem has a role to play. There’s not a silver bullet. We have to make sure every stage of someone’s career funnel is supportive to them staying here and becoming the best in the world.

What do you think needs to happen in order to fix this?

From a Cohere perspective, we just have to keep building great products. Being a winner in our class of products means that we can compete with American salaries and keep talent here. We just have to win in our product categories, so we can keep working on ambitious things that keep young people here, so our early engineers, perhaps can have an exit and they can then also support the ecosystem. Then there’s a reason for companies to start here because there are customers here, whether that’s Cohere or whether that’s companies started by others in the ecosystem.

Cohere is reportedly raising $500 million USD at a $5 billion valuation. Yet you have the same concerns as your fellow vantage points panellist Joella Almeida (MedEssist) in terms of talent and the ability to keep it in Canada. Why does Canada suck at retaining talent?

It’s certainly a nuanced issue. I’ve been thinking a lot about that. I think it goes back to thinking about someone’s career as a funnel. For them to grow and get to the next stage of whatever they’re doing, whether they’re an engineer or an entrepreneur, there’s just no later stage funnel support in our ecosystem.

It’s harder to find ambitious VC dollars, it’s hard to find very experienced senior engineers to learn from. We honestly just need more Shopifys or Coheres to have that financial success to justify more VC dollars. [We need] very ambitious, difficult technical problems that these engineers learn how to solve and they can teach the next generation, for example.

Why do so many companies struggle to find Canadian customers as the first customer or early customers?

I think we may have a little bit of a little brother syndrome [compared] to the States when it comes to our buying culture. Speaking from personal experience, not a lot of companies in Canada took us seriously until we had a few marquee American customers. We needed that stamp of approval down south before we could actually work with a Canadian company.

Why that’s so tragic is we’re here, we’re in Toronto. We would love to work with a Canadian company, not just for revenue. We need the product feedback, and being able to have our engineers onsite across the street on Adelaide, at your offices, seeing the product problems you’re running into, it’s so valuable for us.

[Cohere] not even having a shot in the evaluation phase is a bummer. We are built here. There’s a need for this product. We’re not asking for handouts. Just at least have us in the starting line.

AI companies obviously have high capital, high compute, and high talent requirements. Do you think it’s possible for Cohere to scale globally while being based in Canada, or are you handicapping yourself by building here?

I wouldn’t call it a handicap necessarily. I think being in Canada does give you certain advantages, and if we’re a worthwhile company to have around, we will play to these advantages.

For us, our advantage is capitalizing on being a place where early talent can come and work on ambitious projects and they can get paid a competitive salary. I’d say we’re playing to Canada’s advantage and trying to address the shortcomings by actually having offices elsewhere.

So you think it’s possible for Cohere to continue to scale globally while remaining in Canada?

Yes, 100 percent. I think it’s obviously possible.

Any time a founder working on an AI business or an AI engineer is considering starting their careers in Canada or elsewhere, there’s a big pros and cons list.

I think it’s our job as an ecosystem—not just government, not just VC, but also growth companies, enterprise, community builders—it’s our job to increase that pros list as much as possible, because right now, it’s very hard to stay in Canada if you want to be the best in the world at something.

How are you feeling right now?

Generally, very optimistic. I think the fact that we’re able to consistently win the talent lottery very early on, it’s good. We consistently produce the best in the world when they spawn. I think all these other issues, like the rise of AI and the support from the government, I think we’re able to make this ecosystem more sustainable, and so I’m optimistic that we can fix these issues.

With files from Josh Scott. Feature courtesy Mauricio J Calero for BetaKit.

On May 7, The BetaKit Town Hall provided a pulse check on Canadian innovation, policy, and optimism.

Please enjoy this selection of highlights and insights from the town hall: