VK Ads is a platform that unites all the largest social networks and services in Russia and the CIS with a total reach of over 140 million people. Thanks to VK Ads, advertisers have a single point of access to the Runet audience and a colossal volume of high-quality mobile traffic. The core of the advertising system is an auction, which ensures the effective implementation of bidding strategies chosen by advertisers and ad selection for users.

We are looking for a leader for the advertising verticals team who is interested in deeply delving into the auction mechanics, and growing the effectiveness of advertising products and bidding strategies using machine learning methods.

What we offer:

• The opportunity to implement solutions that affect ~100 million users daily
• An interesting subject area — advertising combines a recommendation system and an auction with many additional mechanisms
• Company technology platforms that simplify runtime implementation

Responsibilities

• Lead the team: hiring, development, performance review
• Decompose business goals and projects into tasks for the team
• Increase advertising effectiveness using ML algorithms
• Propose new solutions and conduct A/B experiments to evaluate hypotheses
• Ensure the stability of ML solutions in a high-load runtime

Requirements

• At least two years of experience leading an ML or development team
• Experience in team development, conducting code reviews
• Practical experience in Java, Python, or C++ development
• Knowledge of classical machine learning methods (linear models, gradient boosting, matrix factorization, NLP methods) and DL
• Understanding of data preprocessing principles and methods, feature engineering, and model evaluation
• Knowledge of algorithms and data structures

Will be a plus

• Experience developing recommendation systems
• Experience with big data (Spark, Hadoop, YTsaurus)