At N.Rich, a leading European ABM platform, I focus on research and development of machine learning models to improve targeting and campaign effectiveness for our customers. My responsibilities include performing statistical analysis, working with large datasets, developing predictive models, and building scalable data pipelines. I collaborate closely with engineering and product teams to ensure our models align with business needs and scale efficiently.

Part of the Dynamic Pricing team at S7 Techlab, a subsidiary of the S7 Airlines group of companies. We operate in a fast-paced environment, focusing on the data-driven optimization of airfare pricing based on variables like cookies, user behavior, and market trends. My primary responsibilities include refining an existing machine learning model and developing a new model for pricing ancillary services.To align our technical strategies with business objectives, I regularly interface with business stakeholders. This cross-functional collaboration ensures that our data science solutions meet real-world needs. On the technical side, I've employed machine learning algorithms to enhance the performance and accuracy of our existing airfare pricing model. I've also used data engineering best practices to create robust data pipelines that feed into our machine learning models.As a result of these efforts, we've seen significant improvements in our current airfare pricing model, making it more accurate and dynamic. Additionally, I developed methodologies for pricing ancillary services, creating a more comprehensive and effective pricing strategy.

Employed adaptive subgradient methods with switches for productive and non-productive steps, successfully applied to the task of maximizing the utility of large computer networks.Analyzed existing methods for solving both the primal and dual tasks for minimizing network utility.Proved the primal-dual nature of mirror descent with switches, enabling an approximate solution for the dual problem based on an approximate solution for the primal.Currently developing methodologies to compare the effectiveness of this approach with other known methods.