TO FIND AN EXPERT ON COMPUTER VISION
One of our old partner’s projects is a popular messenger. At some point, they decided to add virtual masks to it: you know, when you do a selfie and get a crown or long tongue, or cat’s ears drawn on you.
To make it possible, our client needed a C++ and computer vision specialist. An ironclad requirement: a candidate had to have experience in working on this kind of task. The new employee was to jump in and lead the development immediately, with no time for studies.
The problem was that Computer Vision as a field was founded around 2010. There were few CV specialists around, and little to no experienced ones. Across Russia, the number of specialists could be counted not even in the hundreds, but the dozens, and we had to find them all.
MARKET ANALYSIS AND POINT SELECTION
We began by searching for the projects where computer vision is used. Our main field of interest was services that worked with photo and video images of the faces:
social networks and messengers, such as Snapchat or Instagram;
face recognition services like Findface, LUNA system, or Vocord FaceControl;
3D modelling services like Face.DJ;
apps for gender and age recognition based on photos.
After that, we made a list of companies that do something like this and have their own development centres. Then we searched for the employees who work directly with CV. We ended up with 67 people. All we had to do afterwards was to dig up their LinkedIn contacts, go through every resume, and hold dozens of interviews. In the end, the list was cut down to three candidates. Easy as apple pie.
One of them was a perfect fit experience and competencies-wise as he worked on the same technology before. We managed to interest him in the position and to go through the round of interviews. The client offered a good package good conditions, and after the interview, the candidate accepted the offer.
A CV SPECIALIST WHO TAUGHT THE APP TO SEE
To find a CV specialist, we analyzed the market closely and identified every single person who worked on something similar in Russia. The client ended up with the most suitable candidate who managed to build up the whole virtual masks system inside the company.