Arsenii Margasov

Experience

• 

  Sberbank

  Jul 2019 - Jan 2022

  As machine learning engineer improved ML part of SberPM, a comprehensive web platform that enables process mining for large-scale data analysis and optimization.Developed the Python-based process mining library, SberPM, enabling efficient data processing and analysis for process optimization.Successfully transformed SberPM into a Spark/PySpark-based version, optimizing performance and scalability for large-scale data processing and analysis.Conducted extensive research into process mining methods, resulting in the development of novel approaches for data analysis and process optimization.Mentored and provided guidance to two interns, facilitating their professional growth and development. Lihat lebih sedikit During my internship at Sberbank, I had the opportunity to analyze employee headcount models. Through this experience, I developed my skills in data analysis and gained valuable insights into the organization's workforce planning strategies. Specifically, I:Conducted a comprehensive analysis of the employee headcount models, identifying key trends and patterns in the data.Worked with the team to refine and validate headcount forecasting models, resulting in more precise predictions for future workforce needs. Lihat lebih sedikit

  • Machine Learning Engineer

    Jan 2021 - Jan 2022

  • Data Scientist & Analyst

    Sept 2019 - Jan 2020

  • Intern Data Scientist & Analyst

    Jul 2019 - Jan 2019
• 

  Beeline Russia

  Jan 2022 - Jan 2023

  As a data scientist/machine learning engineer, I was responsible for developing response/uplift models to prioritize products for choosing the next best action for clients. Some of my key achievements include:Developing, testing, and deploying an uplift model for returning some type of inactive subscribers, which resulted in reduced retention costs.Developing and testing a model for response to a fintech product, which improved response rates up to 7 times.Developing a model and approach for optimal offering time for subscribers to increase response and reduce communication costs.Extracting valuable insights from the customer's call graph by building embeddings for use in a wide range of models. Lihat lebih sedikit

  • Senior Machine Learning Engineer

    Feb 2023 - Jan 2023

  • Machine Learning Engineer

    Jan 2022 - Jan 2023
• 

  Yandex

  Jan 2023 - now

  - Leading a team of generative answers in Yandex Search, which are based on specified version of YandexGPT for RAG (Retrieval Augmented Generation)- Improved quality/user satisfaction of generative answers up to X% (online search metrics) - Developed LLM for context-based suggestions in dialog (same as in Bing Copilot)

  • Team Lead, Yandex Search, Generative Answers

    Feb 2024 - now

  • Machine Learning Engineer, Yandex Search, Generative answers

    Jan 2023 - Feb 2024