Getting off the ground
In 2017 Davis moved from Boston to Montreal because of its new flourishing reputation as an AI hot spot. He had heard a lot about the deep technology incubator TandemLaunch and decided that was the place to be.
I met Ehsan Saboori at TandemLaunch and that is where we started Deeplite. We hit it off and had very much the same view of AI and deep learning. It is and is going to be the best analytical method for a while, Davis explains.
“We are producing a shovel to the deep learning gold rush”
David and Ehsan also quickly discovered a problem they were both eager to solve. A lot of companies are using tons of computing to do things but very few can afford the computation to do what they want to do for deep learning.
Everybody is going to have this problem. Maybe not the Googles and the Facebooks, but everybody else as they are trying to catch up and compete. So we need to make AI and deep learning more accessible for all of the other companies. Our technology comes up with a new design for neural networks based on the constraints you have. Something that would have taken months for expensive engineers in the past and we automated the process to a few hours or days. We are producing a shovel to the deep learning gold rush.
Collaborating with Zenseact
Through MobilityXlab Deeplite and Zenseact found each other, a meeting that resulted in a proof of concept
Zenseact is obviously very advanced in the field of autonomous driving. They have very advanced algorithms. They gave us a sample model and said “here is something we kind of use in the car, how much better can your technology be?”. So that was perfect for us, that is how we do pretty much all of our proofs of concept. Give us a test and we’ll show you what we can do. We designed a model for pedestrian detection. We took one of their models and made it 12 times smaller and 3 times faster. Those were the lab results and our next step is translating that to live results.
The 3 main advice for startups
Working closely with bigger players in the industry and collaborating within different hubs is a great opportunity for a startup. But everything comes with its challenges. So we asked Davis what he believes startups should pack with them before they set off on their collaborative journeys.
Here’s what he said:
First of all, keep an open mind about the problem you’re solving to actually get the 360 view of the problem faced by the industry. It’s easy to think you know what the problem is and immediately present the technology to solve it. But you have to be mindful of all of the stakeholders. Second of all, you should dare to be straightforward. Being passive won’t get you anywhere. And lastly, you should try to find champions or an evangelist of your solution in the market early on. That’s my advice.