As part on the Search Advertising team, I am working mostly on the advertisement ranking models, but also conduct various related researches• Work on perfecting CTR and other conversion models, achieved significant improvements in advertising revenues• Designed and analyzed over 10 A/B online tests• Enhanced action-to-action attributions and thus improving the model datasets precision• Helped designing ML library for my team to speed up research tasks. Coded typical research pipelines and created automatic report generation• Contribute to internal Lecture Club with presentations on arxiv.org articles, e.g. Google articles on CTR prediction.

As part of RecSys team I conducted researches and developed recommendation models for various bank products• Created user segmentation using investment, transaction and social-demographic features.• Developed neural network-based cross-domain recommendation model from scratch (Pytorch) for personal feed ranking. Collected cross-domain dataset for training.• Conducted a research on the inductive recommendation algorithms which was submitted as my Master Thesis.• Worked on app feed ranking algorithms for concerts/theaters

• Developed Time Series SARIMAX-based model to identify anomalies in various company services• Built insightful user action dashboards using Grafana