Deepomatic: image recognition for operations in the field

Changed on 15/04/2021
The concept behind Deepomatic, a company that emerged from the Inria Paris centre in 2014, is to make visualisation solutions accessible in the field to those who need them the most for their day-to-day activities.
Techniciens montant un échafaudage
© Unsplash / Sol


Founded six years ago, Deepomatic has developed an SaaS artificial intelligence platform that enables their clients to design and implement their own image recognition solutions based on their requirements. The start-up's core business is automating quality control for trade processes and providing assistance to personnel operating in the field.

A start-up created by three enthusiasts

Currently Chief Technology Officer (CTO), Vincent Delaitre first began developing at the age of nine. This was a time when animated films like Toy Story were all the rage, instilling in him a passion for images that would guide him in his studies. Years later, he studied for a PhD on the automatic recognition of human activity in images as part of the WILLOW project team (Inria Paris, ENS-PSL, CNRS). It was during his time there that he met his future associate, Aloïs Brunel, who is now Chief Product Officer (CPO). Aloïs also began programming as a child, only his focus was on security for systems operators and his PhD was on the theoretical foundations of programming languages. The third founder, Augustin Marty (CEO), joined Vincent and Aloïs for the creation of the Deepomatic project after amassing a wide range of professional experience, including the launch of his own start-up MyTab.

After completing his PhD, Vincent was able to make a technology transfer, with support from Inria for the start-up he was planning to set up with Aloïs and Augustin. Through this support, they were able to launch their technology and were put in touch with companies that are still clients of theirs today, primarily via the Prairie institute, prior to joining the incubator Agoranov. The start-up was then granted funding from Bpi France and was able to raise a total of $6.2 million through a fundraiser.

A visual automation solution

Whether it’s detecting meals in company canteens or the wrong types of waste at incineration plants, making driverless vehicles smarter or assisting technicians operating in the field through image recognition, Deepomatic has adapted to the diverse needs of their fifteen or so client companies in France and the USA, including a number of major accounts such as Suez, Compass, Bouygues Telecom and Sanofi.

Deepomatic was listed as a “Strong Performer” in “The Forrester New Wave™: Computer Vision Platform, Q4 2019”. The personalised visual recognition solution for the physical world developed by the company was awarded best solution, and was the only European platform in the rankings.   

The company has around thirty or so employees on the payroll, 40% of whom are women - a figure significantly higher than the current average for tech firms. The three founders are hoping to expand their team and have their sights set on more than 40 employees by the end of 2021.

This year looks set to be a big one for the company: they are aiming to expand into fibre optic connectors, to raise enough money to be able to deploy their solution on a wider scale (particularly in Europe), and to establish themselves as the leading platform when it comes to the use of image recognition for improving compliance monitoring in the field.

Find out more about Deepomatic


An idea, a project... ? Talk to the Transfer, Innovation and Partnerships Department at the Inria Paris research centre