While early adopters, customers, and partners can play a key role in a startup’s success, finding them at home can often be a big challenge for Canadian technology companies.
Count Toronto-based CentML—which was acquired by US chip giant Nvidia last year in a deal that could reportedly top $400 million USD—among this group.
“You don’t even get a chance, you don’t get into the door, even if you have connections.”
Gennady Pekhimenko, Nvidia
CentML developed software designed to help AI models run faster and more efficiently on existing hardware. While the fast-growing company was able to garner interest from domestic venture capital (VC) firms, co-founder and former CEO Gennady Pekhimenko said, aside from Shopify, CentML struggled to find any Canadian businesses interested in working with it, because it was a startup.
Pekhimenko, now senior director of AI software at Nvidia, shared these remarks during a panel on the path from research to revenue hosted by All In Talks on Thursday at Toronto Tech Week.
In the San Francisco Bay Area, CentML had “no problem” finding companies curious about its tech. On its home turf, Pekhimenko said, “You don’t even get a chance, you don’t get into the door, even if you have connections.”
While the money helps, in some cases, Pekhimenko noted that startups just need to get their products into the hands of users to gather feedback and iterate.
Pekhimenko thinks this lack of risk tolerance is a mentality that stems from fear. Fellow panellist Jodi Baxter, vice-president of AI, cybersecurity, and IoT at Telus, agreed.
“We really need to shift that mentality around taking more risks and failing, and how people are investing, whether that’s [Canadian] enterprises or VCs, because that’s how you’re going to grow,” Baxter said.
The best of Toronto Tech Week
BetaKit is Toronto Tech Week’s official media partner. Read all of our coverage here.
Arteria AI co-founder and CEO Shelby Austin, who is building AI software for financial institutions, echoed Pekhimenko’s assertion, noting that many Canadian tech startups had to achieve success in the US first before Canadian companies would adopt their products.
“I don’t know … if you’d ever hear a Canadian company say, ‘We don’t care who has used these guys, we’re going to be the first,’” Austin said during a later panel about Toronto’s AI builders moderated by BetaKit editor-in-chief Douglas Soltys.
Speaking alongside Austin, Foteini Agrafioti, the Royal Bank of Canada (RBC)’s chief science officer and senior vice-president of data and AI, acknowledged that selling to enterprises in Canada is hard—especially for younger, less established companies.
RELATED: Nvidia acquires Canadian AI efficiency startup CentML
“I do think that it’s our responsibility inside Canadian enterprises to learn how to become good buyers of AI, and younger AI from younger AI companies,” Agrafioti said.
RBC has been trying to put its money where its mouth is, including by partnering with Toronto-based large language model developer Cohere last year in what Agrafioti said one might have called “a leap of faith.” She described the arrangement as an “educating” experience for the big bank, which selected Cohere not because it was Canadian but based on the quality of its tech. “It turned out to be quite successful,” she added.
BetaKit is the official media partner of Toronto Tech Week.
All images courtesy All In.