Ievgen Koleinyk

Senior data analyst and analytics team lead with 10+ years of hands-on experience delivering end-to-end analytics, machine learning and forecasting use cases across retail supply chain, CRM/marketing and IT. I build production analytics with Python (pandas, scikit-learn) and SQL — anomaly detection, demand forecasting, customer segmentation, automated ETL and Tableau reporting.

Currently I lead 10 analysts and own a company-wide Customer Data Platform implementation, turning business questions into measurable analytics products and presenting results to technical and non-technical stakeholders. On the AI side I hold 9 certificates from Anthropic and Google covering prompt design, responsible AI use and AI-assisted analysis, and apply them to LLM API integration and AI-assisted data processing in production work. Two master's degrees (engineering + MBA); based in Prague.

BI & Marketing Analytics Manager — Namecheap / Zone3000

Feb 2024 – Present · Prague, CZ / Remote

• Lead a cross-functional team of 10 marketing, operations and web analysts; own the analytics roadmap and mentor junior analysts.
• Designed and deployed a Python anomaly-detection system (Random Forest, scikit-learn) that automatically flags irregularities in marketing and revenue KPIs, replacing manual review of daily reports.
• Manage the end-to-end Customer Data Platform (CDP) implementation, unifying CRM, marketing, product and web data into one analytical layer for segmentation, targeting and attribution.
• Present findings and data-driven recommendations to senior, non-technical stakeholders; set cross-team reporting standards and metric definitions.

CRM Data Analyst — Namecheap / Zone3000

Aug 2021 – Feb 2024 · Dnipro, UA / Remote

• Built ETL pipelines and automated dashboards (SQL, VBA, Tableau), eliminating recurring manual work in weekly and monthly reporting cycles.
• Designed an RFM customer segmentation model used for lifecycle targeting and churn-risk identification across the customer base.
• Supported 100+ email campaigns with data-driven targeting, post-campaign performance analysis and A/B test evaluation.

Support Analyst, IBM Cognos TM1 — DXC Luxoft

Feb 2020 – Aug 2021 · Dnipro, UA / Remote

• Supported an enterprise planning & BI platform (IBM Cognos TM1) for corporate clients: incident resolution, user and access management, client communication.
• Automated permission auditing with Excel, VBA and cloud tooling — reduced manual workload by 20% (~160 working hours freed per year).

Demand Forecast Analyst, Supply Chain — ATB-Market

Oct 2015 – Feb 2020 · Dnipro, UA

• Built short-, mid- and long-term sales and demand forecasting models (Holt-Winters, regression, moving average) for a nationwide store network.
• Developed dynamic SQL + Power Query forecasting and sales-planning templates adopted as the standard across the supply chain department.
• Improved forecast accuracy and reduced stock deviation by 5% (precision 78% → 83%), cutting out-of-stock and overstock losses.

Anomaly detection for marketing & revenue KPIs

Namecheap / Zone3000 · 2024 – present · Python, scikit-learn, Random Forest

A model scores daily KPI series and flags irregularities automatically, so analysts stop eyeballing dozens of daily reports and only look at what actually broke.

Daily traffic stays near 130 all week; Thursday spikes to 800 and is flagged as an anomaly.

Customer Data Platform implementation

Namecheap / Zone3000 · 2024 – present · CDP, SQL, data modelling

Owned end-to-end delivery of a platform that merges four previously separate data domains into a single analytical layer used for segmentation, targeting and attribution.

Four domains unified into one layer: CRM, Marketing, Product, Web analytics.

RFM segmentation engine

Namecheap / Zone3000 · 2021 – 2024 · SQL, Tableau

Recency–Frequency–Monetary model that assigns every customer to a lifecycle segment; used for campaign targeting and churn-risk identification.

Example segment mix: Champions, Loyal, At Risk, New.

Demand forecasting for a national retail chain

ATB-Market · 2015 – 2020 · Holt-Winters, regression, SQL, Power Query

Short-, mid- and long-term forecasting models plus planning templates that became the department standard, cutting both out-of-stock and overstock losses.

Forecast precision: 78% before → 83% after (stock deviation reduced by 5%).

Python · source on github.com/jekosi4ek

ShamelessJobBot

Telegram bot that aggregates and filters job postings; automated collection pipeline with rule- and LLM-assisted relevance scoring.

football_bot

Sports analytics bot and randomizer that collects historical match statistics — an amateur soccer manager.

Programming & Data: Python (pandas, NumPy, scikit-learn), SQL, VBA, Power Query, HTML/CSS, REST APIs, Git

Machine Learning & Statistics: supervised learning, Random Forest, regression, classification, anomaly detection, time-series forecasting (Holt-Winters, moving average, seasonality), hypothesis testing, A/B test analysis, RFM and ABC/XYZ segmentation, IBM SPSS

BI & Visualization: Tableau, QlikView, IBM Cognos TM1, advanced Excel, dashboard & KPI design, data storytelling

Data Engineering: ETL pipelines, data modelling, data quality & validation, Customer Data Platform (CDP), reporting automation

AI & Generative AI: LLM API integration, prompt engineering, AI-assisted data processing, responsible and privacy-aware AI use

Consulting & Leadership: stakeholder management, requirements gathering, team leadership (10 analysts), mentoring, cross-functional communication

Education

• Master of Business Administration (MBA) — Oles Honchar Dnipro National University, 2017–2018
• MSc, Automation and Computer-Integrated Technologies — Dnipro National University of Rail Transport, 2004–2009

Certifications

Languages

• Ukrainian — native
• English — C2 (fluent) Advanced level of instruction — American English Center, May 2019 · Credential ID 016978
• Czech — B1
• Russian — fluent

I started in engineering — a field where a system either works or it does not — and that habit of checking whether something actually holds up carried straight into analytics. An MBA later added the other half: how a business decides, and what a number has to prove before anyone acts on it.

Most of my work sits between those two. Rather than producing reports on request, I try to build the thing that makes the report unnecessary: a model that flags the problem, a pipeline that refreshes itself, a metric definition the whole company agrees on. The anomaly-detection system above exists because four people were reading the same dashboards every morning.

I lead a team of 10 and spend a good share of my week on mentoring — mostly on how to structure a question before touching the data, and how to present a result to someone who will never open a SQL client.

Outside work: cycling, football, and a long-suffering Manchester United habit.

Email: evgenmk87@gmail.com

LinkedIn: linkedin.com/in/ievgen-k-425b06196

GitHub: github.com/jekosi4ek

Location: Prague, Czechia

Download CV (PDF)