Vitaly Kovalev

Work on offline retail analytic and recommendation big data engine. Retail customers behavior research. Description and prediction data analysis.

R&D: combining traditional business data (payment transactions, CRM data, etc) and social network data (Twitter, Facebook, Foursquare) with the help of Hadoop, Storm, Kafka and their ecosystems.

Creating web-spiders for social network data grabbing. Design and development of large-scale systems for analyzing social network data (hundreds of millions of users). Data preprocessing, filtering, recovery of inaccessible information, social relations analysis. Building data crunching systems to process users from multiple social networks. Results were used to create advanced client scoring models for banking. Using distributed systems like Hadoop for massive data processing.

Using data mining tools for a wide range of problems. Custom data mining development. Pattern searching and visualization. Collaborative algorithms development. Large-scale data processing. Creating parallel, high-performance algorithms. Mathematical predictive model analysis, adjustment, and application.