Krystian Korzec

Data Scientist | Decision Optimization, Bayesian Modeling & Causal Inference

đź“Ą Download PDF Version

Contact Information

Professional Summary

Quantitative Data Scientist with over 8 years of experience bridging rigorous statistical methodology with production-grade software engineering. Specialized in Decision Optimization under Uncertainty, Bayesian Statistics, Causal Inference, and A/B Experimentation. Proven track record of architecting end-to-end analytical solutions—from linear programming and hierarchical Bayesian modeling to containerized, production-ready web applications (PyShiny) and scalable GCP data pipelines. Adept at transforming ambiguous business problems into scientifically grounded algorithms that drive measurable margin and operational improvements.

Key Competencies & Tech Stack

  • Machine Learning & Predictive Modeling: Gradient Boosting (XGBoost, CatBoost), Scikit-Learn, Hierarchical Bayesian Modeling, Model Explainability (SHAP).
  • Experimentation & Causal Inference: Online A/B Testing Strategy, Experimental Design, Synthetic Controls, Causal Impact Evaluation.
  • Programming & Data Engineering: Advanced SQL (BigQuery, PostgreSQL), Python (Polars, Pandas, PyMC, Pyomo, Scikit-learn, PyShiny), R, PySpark.
  • Cloud, Infrastructure & MLOps: GCP (BigQuery, Airflow), Docker, CI/CD, Git, AWS, Snowflake.

Employment History

Data Scientist | Allegro

2024 Jul – Present | Warsaw, Poland

  • Formulated and deployed hierarchical Bayesian predictive models and linear programming workflows (Project Charon) for real-time customer service contact allocation under operational uncertainty.
  • Designed and statistically evaluated online A/B experimentation frameworks for the Best Price Guarantee program, quantifying causal impacts across millions of daily seller offer updates.
  • Maintained and optimized production machine learning pricing models to estimate price elasticity and refine automated product selection algorithms.
  • Created in Data Studio a variety of dashboards for model performance diagnostics and A/B tests results overview.
  • Partnered closely with internal product and business stakeholders to translate operational needs into actionable Data & AI solutions.

Tech Stack: Python, GCP (BigQuery, Airflow), PyMC, Pyomo, GitHub, SQL.

Data Scientist | Allianz

2023 Jul – 2024 Jul | Warsaw, Poland

  • Developed and maintained data-intensive statistical applications and analytical pipelines using Python (PySpark, Pandas).
  • Utilized Amazon EMR Studio to perform large-scale data processing tasks for actuarial and analytical workflows.
  • Designed and deployed orchestration logic using open-source Dagster to govern data pipelines, integrating automated freshness monitoring rules and failure notifications.
  • Upgraded and maintained a live Flask API serving risk scoring models, implementing new features and endpoints to support downstream application requirements.
  • Tech Stack: Python, AWS, GitHub, Dagster, SQL

Data Science Analyst | Accenture

2022 Nov – 2023 Jul | Warsaw, Poland

  • Delivered analytical solutions solving complex allocation problems using Integer Programming and Supply Chain Optimization techniques.
  • Spearheaded the refactoring of legacy statistical R code into scalable PySpark/R solutions in Azure Databricks, drastically reducing execution time and enforcing engineering rigor.
  • Tech Stack: Python, GCP, Databricks, GitHub, PySpark.

Data Scientist | Nowa Era

2021 Jan – 2022 Sep | Warsaw, Poland

  • Built predictive models for early detection of at-risk users, transitioning the analytical focus towards proactive intervention strategies.
  • Architected analytical infrastructure and end-to-end data pipelines leveraging AWS and Snowflake.
  • Tech Stack: Python, R, Snowflake, GitHub, Docker, AWS.

Data Scientist | PKO BP

2018 Jul – 2021 Jan | Warsaw, Poland (Promoted to Data Scientist following a 6-month Data Science Internship)

  • Utilized spatial data and ML model (XGBoost) to build Automated Valuation Models (AVM) for real estate pricing prediction.
  • Technical owner of monthly credit risk reporting process.
  • Tech Stack: Python, Hadoop, SQL.

Education

Warsaw School of Economics (SGH)

Master’s Degree in Quantitative Methods in Economics and Information Systems (MIESI) - Master Thesis: Spatio-Temporal Bayesian Model of PM2.5 and PM10 Air Pollution

University of Warsaw (MIMUW)

Completed 2 Years of B.Sc. Coursework in Pure Mathematics - Built a rigorous theoretical foundation in probability theory, real analysis, and linear algebra supporting advanced statistical modeling.