Вероніка

VERONIKA LAKIZA
Data Analyst

CONTACT PROFILE
[відкрити контакти](див. вище в блоці «контактна інформація») Recent Graduate in Artificial Intelligence Systems, passionate
[відкрити контакти](див. вище в блоці «контактна інформація») about machine learning, statistical analysis, and data-driven
insights. Experienced in building end-to-end data pipelines,
Lviv, Ukraine
applying ML algorithms, and turning raw data into clear,
actionable findings - from data cleaning and EDA to anomaly
EDUCATION detection on datasets. Also bring leadership experience
University (SoftServe course), having led collaborative teams to deliver
2022 - 2026
project outcomes.
B.A. in Lviv Polytechnic National
University
Lyceum
PROJECTS
2019 - 2022 Personal Finance & Budget Analysis Dashboard
Ivano-Frankivsk Physics and Built a multi-year budget tracker in Google Sheets from raw
Technology Boarding Lyceum income/expense data. Created pivot-table breakdowns by
category, trend charts over time, and analyzed spending patterns
SKILLS by day of week and month. Used it to identify top expense

SQL (SQLite) drivers, seasonal spending spikes around holidays, and track

Excel/Google Sheets savings rate over time. (Private data, not published.)

Tableau/Power BI Transaction Anomaly Detection
Python Built an anomaly-detection pipeline for 1M payment transactions
Pandas with no labeled data, for a fintech hackathon. Investigated
NumPy suspicious leads manually first — confirming each only when it
scikit-learn showed a clear, concentrated pattern (like a bank ID with zero
Git variation in processing time, or refunds consistently exceeding
Statistics the original amount) rather than normal variance — then cross-
Linear Algebra checked findings with an IsolationForest model. Identified 4
Communication skills clusters of synthetic/corrupted data (0.46% of rows), while
Problem-solving skills correctly ruling out a real provider outage that looked suspicious
Organization and time on the surface.
management skills
Kickstarter SQL Analysis
Analyzed ~380K Kickstarter campaigns using pure SQL (SQLite) —
CERTIFICAT from raw CSV import through data cleaning to analysis. Built
Certificate in Business Analysis cleaning rules to handle mixed currencies, invalid dates, and
ambiguous project states, then compared success rates by
LANGUAGES category, funding-goal size, and launch year. Found that success

Ukrainian - Native rate varies significantly by category (65% down to 24%) and

English - B1+ consistently drops as funding goals increase