Cameron Mills, Chief Operating Officer and Co-founder of Terrene and his co-founders Kashayar Pourdeilami, and Francois le Roux.
Waterloo has been great for us. It has such an ecosystem of entrepreneurship and tech, plus one of the best universities in the world.

From a Waterloo dorm room to an NYC accelerator, Terrene has developed a machine learning platform allowing anyone to do data analytics

While most university students were cramming for finals, the founders of Terrene were balancing their studies with some considerable extra-curriculars. The University of Waterloo students founded a machine learning start-up during their second year of university. “Starting a business when you’re 19 isn’t exactly what school prepares you for,” says Cameron Mills, Chief Operating Officer and Co-founder of Terrene.

Mills and his co-founders Kashayar Pourdeilami, and Francois le Roux created Terrene following their mutual interest in data science. Terrene is an auto-machine learning engine, an end-to-end platform that cleans data, builds models and creates dashboards to give predictions and insights. After all, “Data analytics is useless without insights,” Mills says.

While an increasing number of companies are offering machine learning-powered services today, Terrene’s level of customization sets them apart. “The platform generates a custom machine learning model based on the data you have,” explains Mills, “That’s the difference between the pre-built and the unique model, we make sure customers get the best results for their data every time.”

While analyzing large amounts of data can be difficult or intimidating, Mills says Terrene is accessible for all kinds of businesspeople. “Someone with the comfort level to use Excel can use our platform, but to achieve a much more powerful level of analytics,” he explains.

One of the sectors they have found a growing userbase in is Ontario’s manufacturing sector. “Many automotive and industrial companies have a lot of data but not many data scientists,” says Mills. They have provided manufacturers with demand forecasting to help with inventory and predictive maintenance to detect when machines will need repair before they break.

“Is it Skynet?”

Mills says he sometimes runs into misconceptions about machine learning, simply because people don’t have any experience with it. “They will ask us ‘Is it SkyNet?’” He laughs. “I don’t blame them. Of course, a computer that learns on its own and makes decisions for you, it’s a daunting thing.”

He finds that people are curious about AI, “Knowledge about AI is getting better, 10 years ago I didn’t know what machine learning was, but people still have worries,” he explains.

“There’s lots of articles about how AI is going to take over the world, but these doomsday scenarios might only be true about smarter-than-human level artificial intelligence,” he explains. “AI today is not that advanced, it is really good at answering specific questions and doing specific tasks, but it’s not capable of making its own decisions or figuring new things out unprompted.”

A more practical concern is that of data privacy and cybersecurity, which Mills takes seriously noting recent high-profile breaches in finance and social media. “Anonymizing is the first step,” Mills says, “Machine learning doesn’t care between names and numbers, so all data we use is anonymous.” Secondly, they run their machine learning models on the client’s servers, in order to avoid moving the data and minimizing exposure to breaches. “There’s no excuse not to take this seriously,” he says.

Dorm rooms to boardrooms

Founding a start-up on top of being a full-time student isn’t easy, but the Velocity program at the University of Waterloo helped. Mills had two internship placements at the entrepreneurship program and credits it with developing his interest in start-ups. “Velocity gave us a better understanding of what a start-up is, and gave us legitimacy as well,” Mills says. “There is a big difference when you go from a bunch of kids sitting in a dorm room, to a real office.”

From the incubator environment at Waterloo, Terrene next sought out the Techstars accelerator program in New York City in 2017. “We didn’t know what we didn’t know about business,” says Mills. “Accelerators tear apart your business and help rebuild it from scratch.” This meant three months of intense 9:00 am to midnight days. Techstars has recently brought their accelerator program to Toronto.

Today, Terrene has a home at Communitech’s new Data Hub. “Waterloo has been great for us,” Mills says, “It has such an ecosystem of entrepreneurship and tech, plus one of the best universities in the world. The incubators and accelerators here have really supported us. The most important thing is being constantly surrounded by people as interested as you are in entrepreneurship.”

Terrene has also found Ontario to be a great location for growing their business and finding new customers. “Ontario is a great place to be, especially, for the automotive industry,” Mills explains, “Plus, with Michigan only a four-hour drive away we also have access to all the Detroit automakers”

Terrene has an ambitious year of growth ahead of it as it has just launched a desktop SaaS version of its machine-learning platform. “This will be even more affordable and more accessible,” says Mills, “We want to get our product into the hands of more people so that they can use machine learning too.” Mills notes with the company’s growth they are currently hiring, “We’re ready to grow our team.”