Seeking opportunities for professional growth in data analysis and visualization.
    •  
  • website icon
    •  
    •  
Ru

Data analysis

Skills

  • Python

    pandas, seaborn, numpy, scipy, statsmodels, matplotlib, plotly, sqlalchemy, folium, IPython, os, phik, sklearn

  • PostgreSQL

    Aggregation, grouping, JOIN, subqueries, CTE, window functions

  • Statistical data analysis

    Student's T-test, Mann-Whitney U-test, Fischer Z-test

  • Cohort analysis, unit economics

    LTV, CAC, ROI, ARPU, CR, Retention Rate

  • Prioritization of hypotheses

    ICE, RICE, analysis of A/B test results

  • Basics of Machine Learning

    EDA, regression, classification, clustering

  • Web analytics

    Google Tag Manager, Google Analytics, Google Search Console, Tilda, CRO, VWO, Yandex.Metrica, Yandex.Webmaster
Work and pet projects

  • Prioritization of hypotheses and analysis of A/B test results ➚

    Python: pandas, seaborn, numpy, scipy.stats.mannwhitneyu, matplotlib

  • Market research ➚ of public food establishments in Moscow

    Python: pandas, matplotlib, seaborn, numpy, scipy.stats.ttest_ind, plotly.graph_objects, folium.Map, Choropleth, Marker, DivIcon, folium.plugins.MarkerCluster

  • Analysis of the subscription ➚ book reading service

    PostgreSQL, Python: pandas, SQLAlchemy

  • Analysis of bank ➚ customer churn

    Python: pandas, numpy, matplotlib, seaborn, phik, scipy.stats.mannwhitneyu, scipy.stats.ttest_ind, statsmodels.stats.proportion.proportions_ztest

  • Investigation of the sales funnel ➚ and the results of the A/A/B experiment

    Python: pandas, matplotlib, seaborn, numpy, scipy.stats.ttest_ind, scipy.stats.mannwhitneyu, statsmodels.stats.proportion.proportions_ztest, plotly.graph_objects

  • Increasing website loading speed ➚ and filtering low-quality traffic

    Tilda, Google Tag Manager, Google Analytics, JavaScript, HTML

  • Changes in the value ➚ of the investment portfolio, S&P500 and RTS indices

    Highcharts, Google Sheets, JavaScript

Tableau

Calculated Field, Parameters, Hierarchy, Table Calculation, WINDOW, LOD, GEO, Actions, Filters, Reference & Trend Line, Forecast
Goal:
  • To identify the factors that most significantly influence the success or failure of startups
Objectives:
  • Detect nonlinear relationships between numerical and categorical metrics that standard Pearson correlation fails to capture
  • Provide users with a tool for conducting their own in-depth analysis of the distribution of results
Process:
  • Calculating the Phik correlation matrix in Python to uncover nonlinear relationships
  • Designing a PostgreSQL → Tableau data model
  • Configuring a filtering and navigation panel using Dynamic Zone Visibility
  • Creating explanatory notes, interactive highlights, and tooltips
  • Publishing and presenting the dashboard ⧉
Solution:
  • Interactive dashboard in Tableau, integrated with PostgreSQL
  • Navigation system for data exploration and drill-down into any metric
Impact:
  • Quantification of venture risks through dependency analysis
  • Identification of critical metrics for strategy optimization
  • Filtering out weak projects based on early indicators
  • Increase in venture portfolio ROI to 20-30%
Goal:
  • To visualize the AI startup landscape and identify investment trends across 14 market categories for the period from February 2025 to February 2026
Objectives:
  • Present 305 startups in a structured format while preserving 5 key parameters for each category
  • Enable users to independently compare categories by funding volume, growth dynamics, and market trend
  • Provide quick access to specific companies and market facts through interactive tooltips
Process:
  • Adapting the Veridion external classification as a data source
  • Designing a visual layout in periodic table format
  • Placing parameters using Map Layers: investment ranking, market trend, annual growth, funding volume, number of companies
  • Configuring tooltips with startup examples and analytical facts for each category
  • Publishing and presenting the dashboard ⧉
Solution:
  • Interactive Tableau dashboard with a periodic table layout
  • Navigation across five parameters through visual element placement
  • Tooltips and bar charts as a second level of detail
Impact:
  • Clear comparison of 14 AI market categories in a single screen
  • Identification of growth leaders: AI for Developers (+320%), AI Agents (+260%), Vertical SaaS (+250%)
  • Capturing the market shift from general-purpose AI to industry-specific AI
  • A ready-made tool for initial screening of investment directions in artificial intelligence
Goal:
  • To visualize the risk-return relationship of asset classes and model investment portfolios based on 25 years of historical data (2001–2025)
Objectives:
  • Illustrate the behavior of individual assets and diversified portfolios on the risk-return scatter plot
  • Enable users to evaluate annual returns for any portfolio or asset over a selected period
  • Provide filters by analysis period and asset classes
  • Implement a "Quadrant View" mode to categorize portfolios based on target return and risk tolerance
Process:
  • Collecting and calculating 25 years of historical data on asset returns in USD
  • Creating a scatter plot with three types of objects: diamonds (asset classes), lines (portfolios of two assets), and grey dots (portfolios of three assets in 5% increments)
  • Developing tooltips showing annual returns for each point and diamond
  • Integrating filters by analysis period and asset classes
  • Configuring Dynamic Zone Visibility for "Quadrant View" mode: when enabled, additional filters for target return and risk tolerance appear
  • Publishing and presenting the dashboard ⧉
Solution:
  • An interactive Tableau dashboard based on a scatter plot with three levels of portfolio detail
  • Visual tooltips showing annual returns as a second level of analysis for any point
  • "Quadrant View" mode with dynamically appearing filters to assess portfolio relevance based on the investor's specified parameters
Impact:
  • Visual confirmation of the fundamental relationship between risk and return based on real 25-year data
  • Demonstration of the diversification effect: portfolios consisting of two and three assets visually shift toward the zone with the best risk-return ratio
  • A tool for selecting a model portfolio tailored to an investor’s individual risk profile
  • A ready-to-use framework for client consultations on asset allocation
Goal:
  • To evaluate the effectiveness of ad placement filtering in the Yandex Advertising Network
Objectives:
  • Compare the trends in key metrics before and after ad placement filtering
  • Visualize the distribution of metrics by device type and ad placement quality
  • Set up monitoring of filtering effectiveness with weekly updates
Process:
  • Preparing and cleaning advertising campaign data from Yandex.Direct
  • Designing a dashboard by device type (desktop, mobile, tablet) and ad placement type
  • Configuring interactive highlights
  • Customizing filters by ad placement, detail, and time period
  • Publishing and presenting the dashboard ⧉
Solution:
  • An interactive dashboard in Tableau for monitoring the effectiveness of ad placement filtering by device and platform
  • A system of highlights by device type as a tool for quick comparison of segments
  • Cross-filters for independent analysis of combinations of platforms and periods
Impact:
  • Evidence supporting of the need for filtering: irrelevant sites lower ROAS
  • A tool for regular audits of ad placement sites
  • A methodology has been established: the list of “bad” sites requires moderation at least twice a year
  • 10–15% of the budget should be allocated to ongoing testing of placements on mobile devices, given their potentially high ROAS

Contacts

Alexander Slobodskoi
BI Analyst

Hello!


I'm a bridge between business challenges and data-driven solutions.


Focusing on product growth and marketing efficiency in financial services. I help businesses reduce churn rates, optimize marketing funnels, and identify segments with the highest lifetime value (LTV). Understanding business needs, asking the right questions, and translating requirements into actionable insights - that's what empowers teams to make informed, effective decisions.


Key areas of expertise:


• Ad-hoc analytics: extracting hidden insights to solve immediate business problems (e.g., identified a 15% drop in ROAS across ad networks)


• Data analysis: leveraging SQL and Python to extract, clean, and analyze datasets for 3 cross-functional teams


• Data visualization: building 40+ interactive dashboards in Tableau and Superset for daily operational and strategic monitoring


• Process optimization: automating reporting to streamline operations, saving the team 15+ hours of manual work per week


Currently seeking new opportunities to leverage BI analytical expertise to drive growth. Let’s connect and discuss how your team can reach its goals through data-driven insights!

    •  
  • website icon
    •  
    •