Дмитро

DMYTRO TAIPS
DATA ANALYST
[open contact info](look above in the "contact info" section) • [open contact info](look above in the "contact info" section) • github.com/Taips228 • LinkedIn • Ukraine - Mukachevo
SUMMARY
Data Analyst with a Computer Engineering background and hands-on experience in fraud detection, A/B
testing, and marketplace & partner-profitability analytics. Owns the full analytics workflow independently
— from raw SQL and Python to BI dashboards (Tableau, Metabase) and deployed tools, delivering working
solutions, not just exploratory notebooks.

SKILLS
SQL: Advanced Queries, Window Functions, CTEs, Query Optimization
Databases: MySQL, PostgreSQL, SQLite, Supabase, MongoDB
Python: Pandas, NumPy, Matplotlib, Seaborn, EDA, scikit-learn, XGBoost, SHAP, SMOTE
BI & Stats: Tableau, Metabase, Statistics, A/B Testing, Google Analytics
Dev & Tools: FastAPI, Streamlit, REST API, Next.js (TypeScript), Git, Docker, CI/CD, Prompt
Engineering
Languages: English (B2), Ukrainian (Native)
EXPERIENCE & PROJECTS
Self-employed 2025 — Present
CPA Analytics Dashboard | Python + Google Sheets
Unified 6 disconnected sources (CRM, Keitaro, Affise, GA, tariff tables) into one model, resolving
inconsistent partner identifiers to a single Partner ID key
Built 22 profitability metrics (conversion, EPL, ROI, margin) with last-touch attribution; shipped a
filterable 18-partner dashboard on a documented, auditable pipeline
A/B Test Analysis — Feature Launch Decision | Python + Google Sheets
Analyzed an A/B test (~25,000 users); caught a sample-ratio mismatch and segmented by platform
to expose opposite effects (iOS +13% vs Android −7%), recommending against global launch
Validated skewed revenue with bootstrap CIs and winsorization; delivered a go/no-go write-up with
prioritized next steps Link
"Забирай за дякую" — Marketplace Platform (Team Project) | SQL + PostgreSQL + Metabase
Owned the product-analytics layer in a dev team: designed the event-tracking schema, North Star
metric (Item Success Rate) and KPIs (WAU/MAU, median time-to-give)
Built funnel and cohort analyses in SQL and a Metabase dashboard; tested hypotheses (e.g., 3+
photos convert higher) to pinpoint funnel drop-off
Fraud Detection System | Python + scikit-learn + Streamlit
Built an ML fraud-detection system on 284,807 transactions, comparing 4 models — best: Random
Forest, F1 0.85 / ROC-AUC 0.97
Applied SMOTE for class imbalance and SHAP for per-prediction explainability; deployed a
Streamlit + SQLite app for real-time inference
Voxora.ai | FastAPI + Next.js + Supabase
Independently built the full frontend and contributed to the backend; engineered the AI onboarding
pipeline and character-card prompt system defining core personality and output quality
Laboratory Intern | Flex Jun–Jul 2024
Collected and logged experimental data; ran structured quality checks to maintain accuracy and
catch measurement inconsistencies
EDUCATION
Uzhhorod National University — BSc, Computer Engineering 2022 — 2026 (Expected)
Certificate in Data Analytics, Mate Academy Feb 2026 — View Certificate