DarwinAI: Leveraging Ontario’s talent to make AI easy

AI has been around since the 50’s, and yet we are only now beginning to realize the power of this complex technology, as it sweeps across industries, transforming virtually every business in its path. If there is one place AI is best-suited for, its advanced manufacturing, where it holds the key to bringing the factories of the future to life. But despite the technology’s ubiquity, AI can be intimidating, making it hard for some businesses to adopt industry-leading processes on their own.

Enter DarwinAI, a Waterloo, Ontario based company that is revolutionizing how we build and understand deep learning and neural networks. DarwinAI was founded by Dr. Alexander Wong and Dr. Mohammad Javad Shafiee after seeing how challenging neural networks can be for the average business to build, maintain and understand.

“Even when they are built, neural networks are very difficult to deploy and explain. You’ll often hear about the ‘black box problem’ which is a way describing the opaque nature of AI; it can do all these wonderful things, but we don’t know how it reaches its conclusions,” says Darwin CEO Sheldon Fernandez.

“Our technology uses ‘AI to build AI’, to make neural networks both smaller and explainable. This can be especially powerful when you want to put deep learning on edge-based devices such as phones, TV, watches, and cars.

From the Lab to the Line

Darwin has taken their technologies to the road through their work with automotive companies. “We work with a lot of automotive companies in Europe and Asia including Audi,” says Fernandez.

“They use deep learning for ‘perception’, which is very powerful when it comes to visual understanding. An autonomous vehicle needs to survey a scene in front of it and understand what’s going on. It’s important that they be able to do that quickly. Our technology is used to increase the speed of the neural networks so the car can respond in real time.”

Darwin’s work with Audi, which was showcased in a paper at last year’s British Machine Vision Conference, resulted in a a 90 per cent reduction in the number of hours their engineers spent building and training such networks.

Darwin’s platform has since been used to deliver similar results for large enterprises in aerospace, industrial and consumer electronics.

New Year, New Growth

With Darwin’s current Waterloo offices, they know there is room to expand both in and out of the province. “After our next round of fundraising, we’ll likely explore opening a second office in Toronto,” says Fernandez. From there, Darwin will set its sights on going global. “At some point we probably do want to expand internationally, given that the majority of our clients are in Europe, Asia and the United States.”

In the meantime, their New Year’s resolution is to grow their reach across Ontario.

Diving in to Waterloo’s Talent Pool

When asked what the biggest benefit of operating in Ontario is, Fernandez was boastful about the province’s tech talent – particularly from his alumni. “I went to the University of Waterloo, as did most of the team. When it comes to AI talent, Waterloo and Ontario in general are quite exceptional.”

In Ontario, this high-quality talent often comes at a competitive rate. “The amount of runway that we get with say a million dollars is 40 to 50 per cent more than what you would get if you located your company in San Francisco. Economically, that’s a real benefit.”

Darwin also credits the provincial and federal governments for their success. “[Programs like] the Industrial Research Assistance Program, Scientific Research Program and the Experimental Development Program make it extremely easy to run a tech company in Ontario. Not to mention how accepting Canada’s been with foreign workers and immigrants who contribute exceptional skills to the workforce.”

A Decade in the Making

Given the potential of AI to transform our way of life, the need to understand the technology will be paramount in the years ahead.

As such, Fernandez believes that explainability is one of the primary challenges when it comes to people embracing AI – and Darwin has the solution. “AI is going to continue to grow at a geometric pace over the next couple of years and we offer really differentiated technology in the space. We’ve been working on this for over a decade – now is the time to realize its potential.”