Олександр

OLEKSANDR ROMANIUK
DATA ANALYST
Lviv, Ukraine • [open contact info](look above in the "contact info" section) • [open contact info](look above in the "contact info" section)

SUMMARY
Data Analyst with a strong foundation in systems analysis, statistics, and the full data lifecycle: from strategic business
analysis (SWOT, Porter's Five Forces) and process modeling (BPMN/UML) to relational database design (SQL, ERD) and end-
to-end machine learning pipelines. Trained through the Lviv IT Cluster program. Looking to apply analytical rigor and hands-
on Python/SQL skills to a Data Analyst role.

TECHNICAL SKILLS
Languages & Querying: Python (pandas, numpy, scikit-learn, XGBoost), SQL, MS SQL Server
Data & BI: Excel, Power BI, Database Design (ERD), Data Warehouse (DWH) / ETL
Machine Learning: Random Forest, XGBoost, SMOTE, feature engineering, model evaluation (F1, ROC-AUC, MCC)

ANALYTICAL METHODS
Business & Process Analysis: SWOT, Porter's Five Forces, Business Motivation Model (BMM), BPMN 2.0 (AS-IS / TO-BE)
Systems & Decision Analysis: UML (Use Case, Class, Activity, Sequence), Analytic Hierarchy Process (AHP/MAI), MCDM

EDUCATION
B.Sc. in Systems Analysis: Business Analysis & Data Science 2022 –2026
Lviv Polytechnic National University (in partnership with Lviv IT Cluster)

FEATURED PROJECT — BACHELOR'S THESIS

Modeling and Analysis of Inauthentic Behavior Criteria for Social Media Users Feb 2026 – Jun 2026
End-to-end decision-support system combining a multi-criteria weighting method with a machine learning classifier to detect bots,
fake accounts, and coordinated networks on social media.
● Designed a 4-category, 24-feature criteria system (profile, behavioral, network, content) and weighted it using the
Analytic Hierarchy Process (AHP), achieving a consistency ratio CR = 0.0115 (< 0.10 threshold).
● Built and trained an ensemble classifier (Random Forest + XGBoost with SMOTE class balancing) in Python, reaching
ROC-AUC = 0.932, F1 = 0.80, MCC = 0.692 on the public Twitter Human-Bots Dataset (37,438 accounts).
● Modeled the business process in BPMN 2.0 (AS-IS/TO-BE) and documented system architecture through 4 UML
diagrams (Use Case, Class, Activity, Sequence) plus an ER-diagram covering 5 entities.
● Ran a 6-scenario sensitivity analysis on the AHP weights, proving the weighted model outperforms uniform
weighting by 1.1 ROC-AUC points and validating the robustness of the criteria system.
● Delivered a fully reproducible, modular codebase (AHPWeighter, FeatureEngineer, BotClassifier classes) with fixed
random seeds and documentation.

OTHER ACADEMIC PROJECTS

Database Systems: Designing a Relational Database for a Dealer Network Nov 2025 – Dec 2025
● Developed the logical structure of a relational database, performed full normalization, and implemented complex
queries based on relational algebra.
● Prepared technical documentation in accordance with international standards.

Data Warehouse Design: Wine Retail Network Nov 2024
● Designed a data warehouse (DWH) architecture and implemented ETL packages to automate data collection and
transformation.
● Developed algorithms for automatic document classification to speed up analytical reporting.

Business Data Analysis: Strategic Analysis of ATB-Market May 2025
● Conducted quantitative and strategic market-position analysis and developed a Business Motivation Model (BMM).
● Mapped business processes using BPMN 2.0, identified bottlenecks, and developed a TO-BE optimization plan.