Deeplite raises $6M seed to deploy ML on edge with fewer compute resources

One of the issues with deploying a machine learning application is that it tends to be expensive and highly compute intensive. Deeplite, a startup based in Montreal, wants to change that by providing a path to reduce the overall immensity of the representation, enable it to run on hardware with far fewer resources.

Today, the company announced a$ 6 million seed financing. Boston-based venture capital firm PJC produced the round with help from Innospark Ventures, Differential Ventures and Smart Global Holdings. Somel Investments, BDC Capital and Desjardins Capital too participated.

Nick Romano, CEO and co-founder at Deeplite, says that the company aims to take complex deep neural networks that require a lot of compute influence to run, tend to use up a lot of recall, and can devour artilleries at a rapid pace, and help them run more efficiently with fewer resources.

” Our programme allows us to change those models into a new species point to be able to deploy it into constrained hardware at the leading edge ,” Romano clarified. Those devices could be as big as a cell phone, a drone or even a Raspberry Pi, meaning that makes could deploy AI in ways that simply wouldn’t be possible in most cases right now.

The company has created a produce called Neutrino that lets you specify how you want to deploy your pose and how much you can compress it to reduce the overall sizing and the resources required to run it in yield. The idea is to run a machine learning application on an extremely tiny footprint.

Davis Sawyer, director commodity patrolman and co-founder, says that the company’s solution comes into play after the pose has been improved, civilized and is ready for production. Users quantity the simulation and the data set and then they can decide how to build a smaller model. That could involve reducing the accuracy a bit if there is a tolerance for that, but principally it involves selecting a level of constriction — how much smaller you can attain the model.

” Compression increases the size of the model so that you can deploy it on a much cheaper processor. We’re talking in some cases going from 200 megabytes down to on 11 megabytes or from 50 megabytes to 100 kilobytes ,” Davis explained.

How artificial intelligence will be used in 2021

Rob May, who is leading the investment for PJC, says that he was impressed with the team and the technology the startup is trying to build.

” Deploying AI, especially deep study, on resource-constrained machines, is a broad challenge in the industry with scarce AI talent and know-how available. Deeplite’s automated software solution will create significant economic benefit as Edge AI continues to grow as a major compute paradigm ,” May said in a statement.

The idea for the company has roots in the TandemLaunch incubator in Montreal. It launched officially as a company in mid-2 019 and today has 15 employees with plans to doubled that by the end of this year. As it constructs the company, Romano says the founders are focused on building a diverse and inclusive organization.

” We’ve got a strategy that’s going to find us the right people, but do it in a way that is absolutely diverse and all-inclusive. That’s all part of the DNA of the organization ,” he said.

When it’s possible to return to work, the project is to have parts in Montreal and Toronto that act as centres for employees, but there won’t be any requirement to come into the office.

” We’ve already discussed that the general approach is going to be that people can come and go as they satisfy, and we don’t think we will need as big an office footprint as we may have had in the past. People will have the option to work remotely and virtually as they see fit ,” Romano said.

To solve all the small things, look to daily Little AI

Read more:

No Luck
No prize
Get Software
Free E-Book
Missed Out
No Prize
No luck today
Free eCourse
No prize
Enter Our Draw
Get your chance to win a prize!
Enter your email address and spin the wheel. This is your chance to win amazing discounts!
Our in-house rules:
  • One game per user
  • Cheaters will be disqualified.