The Top-10 Russian Artificial Intelligence Startups
There are some frequent stereotypes you’ll hear about Russian men, mainly that they’re all a bunch of alcoholic
football soccer fans who die early because of their love for vodka, leaving behind some of the most beautiful women on the planet who outlive them by ten years on average. While the stereotypes have some serious merit, Russia isn’t just about hard-drinking, chasing skirts, and hooliganism. All that vodka fuels some of the world’s brightest minds, like Russia’s fearless leader, Vladimir Putin, who believes that artificial intelligence will be the key to global supremacy. So, what about Russian artificial intelligence technology?
Artificial Intelligence in Russia
As China and the US are fighting for AI supremacy and the EU is scrambling to catch up, Russia’s growing its footprint in AI with a seemingly low-key approach. Hearing less about Russian AI companies doesn’t mean they don’t exist though. The Russian government’s investments into AI have been dwarfed by the billions of dollars China and the US have been spending but this may change with a growing number of public-private partnerships and the involvement of the Russian Ministry of Defense in AI projects. Russia also has a strong venture capital infrastructure in place that has helped grow her “AI industry” up to this point.
Science Guide, a Russian community dedicated to the development of applied science, has released a report on the State of AI in Russia in 2017 which lists 150 Russian artificial intelligence companies. According to the report, most AI startups are working on machine vision (33%) followed by decision-making algorithms (17%) and image/video recognition (14%). Russia seems to get much less coverage on the global stage, so we decided to analyze and cross-check the above-mentioned Science Guide report with Crunchbase to search the Russian AI ecosystem for the most funded startups in Russia. Here’s what we came up with.
The Russians are Watching
Founded in 2012, Moscow startup Visionlabs has raised $5.5 million to develop computer vision solutions for facial recognition, object recognition, and augmented and virtual reality applications. The team created an open-source platform partnering with Facebook and Google in 2016 to help startups launch projects in the field of computer vision worldwide. The collaboration was followed by a Series A funding coming from the venture capital arm of a Russian conglomerate. Since then, Visionlabs has partnered with major banks and corporates to provide their Luna biometric identification platform as a service, incorporating it into customer identification processes. By the end of last year, Sberbank, a leading Russian bank, acquired a 25% stake in Visionlabs to jointly work on building a face, voice, and retina ID platform for its clients. The company is also working with Siberia Airlines to provide personalized services to its business class travelers based on facial recognition.
We first came across RoboCV in our article on “9 Industrial Robots for Your Warehouse”. Founded in 2012, this Moscow startup has raised $3.7 million to develop autonomous vehicle technology for use in logistics and warehousing. Their X-MOTION NG technology is aimed at warehouses and fast-moving consumer goods distribution centers and can be implemented without any major overhaul of existing processes.
Forklift movements are based on predefined routes on a map, and the connected vehicles use these to navigate their surroundings and execute tasks defined on the fly without the need for a human operator. The forklifts communicate with each other to improve efficiency and can manage high traffic areas of both machines and people. The technology is ready for implementation into other types of vehicles as well. RoboCV works with Samsung and Volkswagen among other clients and boasts return on investment periods as low as 2 years for some of their projects.
That’s a Lada Data
Founded in 2012, Moscow startup Double Data has raised $3.2 million to develop big data solutions for banks and financial institutions. Their AI algorithms analyze public information on the web such as social media likes and interests, connections, employment history, and income to help with loan applications and scoring, debt collection, and credit fraud. Double Data has implemented more than 100 projects so far, but has been challenged by VKontakte, a Russian competitor of Facebook this February. VKontakte sued Double Data for unauthorized use and selling of their public profile data and won, but Double Data is planning to appeal against the court decision arguing their freedom to access public information on the internet.
How to Tell if You Look Drunk
Founded in 2011, Miass startup 3DiVi has raised $2.7 million to develop 3D computer vision technologies for gaming platforms, digital signage, and virtual reality applications. Where 3DiVi stands out from the machine vision crowd is their ability to recognize gestures and track full-body skeletal movements in 3D.
Their facial recognition algorithms have been used in diverse scenarios like airport biometric checkpoints, Android augmented reality glasses for law enforcement, and advertisement audience profiling. Their body movement tracking app called Nuitrack comes with their own sensor developed jointly with Orbbec, a specialist in-depth sensors, that aims to replace Microsoft’s Kinect offering. Nuitrack is also compatible with sensors from other providers. The service is subscription-based, includes a software developer toolkit and starts from as little as $20 per annum of $40 for a perpetual license per sensor.
Another Distraction for Russian Drivers
Founded in 2013, Kazan startup RoadAR has raised $2.5 million to develop an augmented reality car navigation app utilizing machine vision. Taking on competitors such as Waze, RoadAR offers real-time updated maps using computer vision to crowdsource map updates, augmented reality when used as a dashcam, real-time notifications of events on the road, and supports voice and text messaging with other users so you can easily figure out which bar to drive to next.
They’ve included audio ads in the app based on geolocation, which are probably annoying but at least keep the app free. It is available from the App Store and from Google Play (under the name Roadly) and has received mixed reviews so far. The team also developed a license plate recognition algorithm that is 98% accurate, and is also selling consultancy services.
The Russian Bots are Coming
Founded in 2014, Perm startup Promobot has raised $2 million to develop an autonomous customer service robot for businesses. According to the developers, Promobot can “work as an administrator, promoter, host, or a museum guide. It can operate in cinemas, museums, shopping malls, business centers, convention centers and exhibition halls, as well as other crowded places.“ They’re not sugarcoating the message, this robot is out to get your job. And it doesn’t even require a uniform.
It also doesn’t show up to work on Monday morning smelling like vodka. Promobot was briefly in the news after “escaping from the lab” in 2016, which was probably just a clever publicity stunt. This was followed by other weird stories, like how it allegedly saved a little girl from a shelf collapsing on her. While the Promobot marketing team keeps grasping for straws, the company’s developers are steadily progressing the robot’s capabilities. Version 3 comes with a scanning module, and the latest (Version 4) is equipped with a card dispenser and can issue security passes to visitors. Promobot is planning to open its first foreign office later this year in the US.
You Must Think in Russian
Founded in 2012, Saint Petersburg startup Cubic Robotics has raised $1.5 million to develop an AI personal assistant. Consisting of a cube-shaped home unit and portable “badge”, it can connect to and control a number of home devices like thermostats, lights or the TV. It also interacts with your smartphone and wearables, suggesting TV programs to watch, measuring calories burned during exercise, or ordering takeout.
Cubic started out with a successful Indiegogo crowdfunding campaign and joins a handful of other startups in the AI personal assistant space. The Cubic team is also developing custom Amazon Alexa skills for businesses like setting up reminders with voice commands and receiving reminders through Facebook Messenger.
Familiar Russian Faces
Founded in 2015, Moscow startup NTechLab has raised $1.5 million to develop their own version of a facial recognition algorithm. Their service not only verifies or identifies faces, but recognizes age, gender, and emotions. The company’s algorithms have received praise in Washington last year, winning the first-ever facial recognition competition devised by the R&D team run by the Director of National Intelligence. The company has 2,000 customers globally including the UK, US, and China, and works with VKontakte as well, the Russian Facebook competitor that sued Double Data. Applications for their technology include public safety, dating, security, banking, retail, entertainment, and events organization.
In Soviet Russia, Video Watch You
Founded in 2009, Moscow startup Synesis has raised $1.4 million to develop a video analytics platform called Kipod that functions like Google Search for video content. Used by law enforcement agencies, governments, and private security organizations, it is able to find human faces, license plates, object features, and behavioral events in massive amounts of footage. What’s really unique about their cloud-based service is its scalability: the company claims it can work with one million users, one million cameras, and search one petabyte (a million gigabytes) worth of videos. The team is now located across 5 countries and has more than 100 client accounts.
The Coldest HR Person on Earth
Founded in 2015, Moscow startup Stafory has raised $1.1 million to develop a hiring robot called Vera. Stafory claims Vera can search job sites and internal databases for candidates, screen them, talk to them, and even hold video interviews recognizing candidates’ emotions.
Vera is currently conducting 50,000 interviews daily for Stafory clients including IKEA, Microsoft, Burger King, and Raiffeisen Bank, and can find top candidates 10 times faster than human HR people. The startup is joining the growing ranks of AI startups in recruiting, freeing up HR personnel for “more value-added activities” and making corporate environments slightly more bearable in the process.
We were puzzled to see that these ten startups have received less than $30 million collectively, yet are operating in the same global markets as more funded startups, achieving recognition and winning international clients. Could it be that all the good talent relocates before their startups get big? This also raises some interesting questions about where a startup should be categorized geographically. Is it defined by the founders’ last names? The investors’ domicile? Or maybe the headquarters location?
Two examples we came across were Cubic Robotics and Synesis, both of which started out in Russia but recently moved their headquarters to the US and Canada respectively. As their origins and investors are clearly Russian, we consider them Russian startups. However, this highlights the importance of location for a startup, particularly when coming from an emerging economy – and leaves us with two important questions. Does moving a startup’s headquarters change its country classification, and could all of Russia’s AI talent be migrating to greener pastures?
Despite what the pundits say, FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) don't give you real exposure to AI. Read about 7 stocks that give you true pure-play exposure to AI in our guide to investing in AI healthcare companies, freely available to Nanalyze Premium subscribers.