Looking for a job

Anton Andreitsev

@antonandreitsev

Senior Machine Learning Engineer at Constructor, building retrieval and reranker models for personalized recommendations with positive A/B test results.

Russian Federation

Message

At Constructor, I build retrieval and reranker models for personalized e-mail recommendations, with positive A/B test results in CTR, CVR, and AOV. I’ve also introduced embedding-based features into the production reranker.

At Alfa-Bank, I built investment propensity models that raised campaign response rates from 0.6% to 2% while cutting spend, and recommenders that increased active clients by 3%. I also taught machine learning at Lomonosov Moscow State University, where I led lectures and seminars for seven years.

Experience

Work history, roles, and key accomplishments

Current

Senior Machine Learning Engineer

Current

Constructor

Feb 2026 - Present (8 months)

Built retrieval and reranker models for personalized e-mail recommendations and introduced embedding-based features into the production reranker.

Machine Learning Personalization Ranking Recommendation Systems A B Testing Embeddings

Machine Learning Engineer

Constructor

May 2024 - Feb 2026 (1 year 9 months)

Improved personalization across search and recommendations, shipped segment-based reranking and propensity models, and used LLM as a judge for dataset labeling.

Machine Learning Personalization Ranking Recommendation Systems Propensity Modeling LLM A B Testing

Lecturer

Lomonosov Moscow State University

Sep 2018 - Jan 2026 (7 years 4 months)

Taught lectures and led seminars for the course 'Introduction to Machine Learning' for seven years.

Teaching Machine Learning

Senior Data Scientist

Alfa-Bank

Oct 2020 - Mar 2023 (2 years 5 months)

Led a team of 3 data scientists, built investment propensity and churn models, and delivered forecasting and wage models. Implemented the team's ML pipeline standard.

Data Science Propensity Modeling Uplift Modeling Churn Modeling Recommendation Systems Forecasting ML Pipelines Python SQL

Data Scientist

Rubbles

Oct 2019 - Oct 2020 (1 year)

Built a daily demand-forecasting model for KFC restaurants and ran R&D on forecast interpretability. Also forecast motor-oil blend characteristics for Gazprom.

Data Science Demand Forecasting Gradient Boosting Model Interpretability Python