Vladislav Morozov, PhD
Data Scientist
Applied data scientist with research-level expertise in causal inference and machine learning. Experienced in informing complex product and business decisions, from experiments to observational data.
- Location
- Barcelona, Spain
- vladislav.v.morozov [at] gmail.com
- Website
- https://vladislav-morozov.eu
- Vladislav Morozov
- GitHub
- vladislav-morozov
Experience
– present
Senior Data Scientist at Glovo
Highlights
- Developing and applying causal methodology to measure the long-term impact of app assets promoting a new Retail category, finding a ~20% category GMV lift per extra early impression
- Building and deploying an ML system to identify Q-Commerce products likely to be out of stock, from data products through production infrastructure.
- Building Global Vendor Categorization and automated catalog filtering systems using LLM-generated product attributes to improve restaurant discovery.
- Advancing experimentation and causal inference practices through the DeliveryHero Experimentation Council and local initiatives.
- Developing evergreen standards and automated tooling for reproducible experimentation and analysis
–
Assistant Professor of Statistics (tenure track) at University of Bonn (Bonn, Germany)
Highlights
- Developed and validated causal inference methods for complex observational data (e.g. 2 top field publications, with Python package and MATLAB implementations)
- Led modernization of data science instruction and launched 3 new courses (causal inference with unobserved heterogeneity; econometrics; simulations in data science) to 90%+ positive evaluations
- Supervised and mentored 15 BSc/MSc students in causal inference and ML, guiding projects and code reviews
–
Research Technician at Universitat Pompeu Fabra (Barcelona, Spain)
Highlights
- Prototyped scalable PySpark/Docker pipelines for novel high-dimensional time series analysis, demonstrating feasibility for large-scale deployment and value of further research
- Contributed to 'Modern Challenges in High-Dimensional Data Analysis' and 'New Frontiers in Econometrics' grants.
–
Doctoral Researcher and Instructor at Universitat Pompeu Fabra (Barcelona, Spain)
Highlights
- Led own research on causal inference for heterogeneous panel data, designing estimators, validating via simulation and large micro datasets, published dissertation 'Essays in Heterogeneous Panel Data Econometrics'
- Taught 40+ sections of data science, econometrics, and machine learning courses (undergraduate to professional), with 4.5+/5 evaluations
- Designed practical labs and exercise sessions: wrote theoretical exercises and created Python/R/Stata materials for econometrics, forecasting (undergraduate/PhD) and NLP for Economics (BSE Data Science Summer School)
–
Junior Antitrust Data Analyst at Federal Antimonopoly Service (Moscow, Russia)
Highlights
- Developed the statistics and implementation for a system to triage consumer complaints about price collusion. By analyzing scanner data, the system increased the share of complaints receiving review from 60% to 90%
- Built and automated dashboards to track competitiveness across 15 key food groups, establishing the authority's first real-time data pipeline from a fragmented data lake
Education
–
PhD in Econometrics from Universit Pompeu Fabra with GPA of Excelente Cum Laude
–
Master of Research in Economics, Finance, and Business from Universit Pompeu Fabra with GPA of 9.3/10
–
Master of Science in Economics from Barcelona School of Economics with GPA of 9.1/10
–
Bachelor in Economics from Moscow State University with GPA of 5.0/5.0
Publications
Inference on Extreme Quantiles of Unobserved Individual Heterogeneity by Econometric Theory
Methods for inference on tails of distributions when only noisy data is available
Unit Averaging for Heterogeneous Panels by Journal of Business and Economic Statistics
Efficient ensemble prediction/estimation of individual forecasts/parameters in heterogeneous data settings
Languages
- English
- Fluency: Native speaker
- Spanish
- Fluency: Native Speaker
- Russian
- Fluency: Native Speaker
- German
- Fluency: Limited working proficiency
- Catalan
- Fluency: Elementary proficiency
Skills
- Programming
- Keywords:
- Classic and Deep Machine Learning, Causal Inference
- Keywords:
- Data and DevOps
- Keywords:
