https://github.com/elissorokin/data-analyst-portfolio
Это репозиторий, в котором я демонстрирую свои навыки, делюсь проектами и отслеживаю прогресс в области анализа данных и Data Science.
https://github.com/elissorokin/data-analyst-portfolio
ab-testing data data-analysis datalense matplotlib numpy pandas plotly portfolio postgresql python scipy seaborn sql statistical-analysis
Last synced: about 1 month ago
JSON representation
Это репозиторий, в котором я демонстрирую свои навыки, делюсь проектами и отслеживаю прогресс в области анализа данных и Data Science.
- Host: GitHub
- URL: https://github.com/elissorokin/data-analyst-portfolio
- Owner: ElisSorokin
- Created: 2025-05-15T16:03:56.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2025-05-31T19:28:40.000Z (9 months ago)
- Last Synced: 2025-06-01T07:31:31.504Z (8 months ago)
- Topics: ab-testing, data, data-analysis, datalense, matplotlib, numpy, pandas, plotly, portfolio, postgresql, python, scipy, seaborn, sql, statistical-analysis
- Homepage: https://github.com/ElisSorokin/Data-Analyst-Portfolio
- Size: 2.03 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Elisei Sorokin - Data Analyst Portfolio
## About Me
Hi! I'm Elisei Sorokin. I have a strong background in mathematics and statistics, and I studied economics and finance at the University of Tor Vergata in Rome. Currently, I’m actively developing in the field of data analysis and applying my knowledge in practice.
Throughout my studies and work, I’ve learned how to work with large and complex datasets, analyze user behavior, build metrics, and test hypotheses using Python and SQL. I’ve also gained experience in conducting statistical and A/B tests to support data-driven decision-making.
In my free time, I explore new tools for data analysis and visualization, and I continuously improve my skills in programming and statistics. I enjoy working both independently and in a team and always strive to uncover valuable insights and apply them to real business problems.
My CV is available in[[pdf](https://github.com/ElisSorokin/PortfolioProjects/blob/main/Sorokin%20Elisei%20(1).pdf)].
This repository showcases my data analysis skills, project work, and learning progress in data analysis and data science.
## Table of Contents
- [About Me](https://github.com/ElisSorokin/Data-Analyst-Portfolio#about-me)
- [Projects](https://github.com/ElisSorokin/Data-Analyst-Portfolio#projects)
- Python & SQL
- [Afisha Perfomance](https://github.com/ElisSorokin/Data-Analyst-Portfolio#afisha-performance--user-activity-and-seasonal-trends-analysis)
- Python
- [Books & Conversion — User Activity Analysis & Interface A/B Test](https://github.com/ElisSorokin/Data-Analyst-Portfolio#books--conversion--user-activity--ab-test-analysis)
- [Recommendation Boost — A/B Test of a New Algorithm](https://github.com/ElisSorokin/Data-Analyst-Portfolio#recommendation-boost--ab-testing-a-new-algorithm)
- SQL
- [Data-Delivery — Analytics of a Food Delivery Service](https://github.com/ElisSorokin/Data-Analyst-Portfolio#data-delivery--analytics-for-a-food-delivery-service)
- [Property Pulse — Real Estate Market Analytics in St. Petersburg and the Region](https://github.com/ElisSorokin/Data-Analyst-Portfolio#property-pulse--real-estate-market-analysis-st-petersburg--region)
- [DarkForest — Player Behavior & In-Game Purchases Analysis](https://github.com/ElisSorokin/Data-Analyst-Portfolio#darkforest--in-game-purchase-behavior-analysis)
- [Education](https://github.com/ElisSorokin/Data-Analyst-Portfolio?tab=readme-ov-file#education)
- [Courses](https://github.com/ElisSorokin/Data-Analyst-Portfolio?tab=readme-ov-file#courses)
- [Contacts](https://github.com/ElisSorokin/Data-Analyst-Portfolio?tab=readme-ov-file#contacts)
## Projects
This section presents data analysis projects with brief descriptions of tasks, tools, and outcomes.
### Afisha Performance — User Activity and Seasonal Trends Analysis
**Code:** [`Afisha_Perfomance.ipynb`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/PT%20(Final)%20-%20Afisha%20Perfomance.ipynb)
[`Afisha_Perfomance.sql`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/SQL%20-%20Afisha%20Perfomance.sql)
[`Afisha_Perfomance Dashboard`](https://datalens.yandex/cnchy35jrh7ky)
**Goal:** Build a dashboard to monitor business KPIs and explore user behavior on the Afisha platform.
**Description:** As a data analyst on the Afisha team, I developed an interactive dashboard in Yandex DataLens and conducted an in-depth behavioral analysis using Python. The analysis focused on user preferences during autumn 2024 and tested hypotheses about mobile vs desktop user behavior.
**Key Contributions:**
- Cleaned and transformed data; engineered new features (revenue, AOV, seasonality)
- Calculated key metrics: revenue, average ticket price, category distribution
- Built a dashboard with filters, segmentation, and KPI dynamics
- Conducted EDA and tested behavioral differences using Mann–Whitney U test
- Identified user behavior trends and seasonal patterns
**Skills:** PostgreSQL (SQL), Python, hypothesis testing, segmentation, metric building, EDA.
**Tools:** Python, Pandas, Numpy, Seaborn, Matplotlib, SciPy, PostgreSQL (SQL), Yandex DataLens.
**Results:**
- Autumn order volume 2.5× higher than in summer
- Mobile users showed higher activity (8.73 vs 6.66 orders, p < 0.0001)
- Most popular: theatre, family, and sports events
- Weekly activity patterns with peaks on Tuesdays and Thursdays
**Recommendations:**
- Invest in mobile development
- Launch seasonal marketing campaigns
- Promote family events and strengthen regional partnerships
---
### Books & Conversion — User Activity & A/B Test Analysis
**Code:** [`Books & Conversion.ipynb`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/PT6%20-%20Books%20%26%20Conversion%20%E2%80%94%20%D0%90%D0%BD%D0%B0%D0%BB%D0%B8%D0%B7%20%D0%BF%D0%BE%D0%BB%D1%8C%D0%B7%D0%BE%D0%B2%D0%B0%D1%82%D0%B5%D0%BB%D1%8C%D1%81%D0%BA%D0%BE%D0%B9%20%D0%B0%D0%BA%D1%82%D0%B8%D0%B2%D0%BD%D0%BE%D1%81%D1%82%D0%B8%20%D0%B8%20AB-%D1%82%D0%B5%D1%81%D1%82%20%D0%B8%D0%BD%D1%82%D0%B5%D1%80%D1%84%D0%B5%D0%B9%D1%81%D0%B0.ipynb)
**Goal:** Analyze user activity on Yandex.Books and evaluate the impact of a new e-commerce UI via A/B testing.
**Description:** This project contains two independent cases:
1 - Exploring user activity by city on Yandex.Books
2 - Evaluating an A/B test for a new interface on the BitMotion Kit e-commerce platform
**Key Contributions:**
- Cleaned datasets, removed duplicates
- Calculated activity metrics by region and prepared test samples
- Used t-test and Mann–Whitney test to compare users in Moscow vs St. Petersburg
- Validated test design (balance, overlap, duration)
- Calculated conversion rates and applied z-test for comparison
- Prepared a business report with actionable insights
**Skills:** tatistical analysis, A/B testing, hypothesis testing, data cleaning
**Tools:** Python, Pandas, SciPy, Statsmodels, PostgreSQL (SQL)
**Results:**
- No statistically significant difference in user activity between cities (p > 0.05)
- Median values were more robust due to skewed time distributions
- A/B test showed a 1.7 pp conversion increase (p = 0.0283)
**Conclusion:** The new interface had a positive effect on conversion and was recommended for rollout.
---
### Recommendation Boost — A/B Testing a New Algorithm
**Code:** [`Recommendation Boost.ipynb`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/PT5%20-%20Recommendation%20Boost%20%E2%80%94%20AB-%D1%82%D0%B5%D1%81%D1%82%20%D0%BD%D0%BE%D0%B2%D0%BE%D0%B3%D0%BE%20%D0%B0%D0%BB%D0%B3%D0%BE%D1%80%D0%B8%D1%82%D0%BC%D0%B0%20%D0%B2%20%D0%BF%D1%80%D0%B8%D0%BB%D0%BE%D0%B6%D0%B5%D0%BD%D0%B8%D0%B8.ipynb)
**Goal:** Evaluate the effect of a new recommendation algorithm on user behavior via a properly designed A/B test.
**Description:** End-to-end A/B test analysis—from calculating the required sample size to drawing business conclusions.
**Key Contributions:**
- Analyzed user registrations and group assignments
- Measured page views per session
- Calculated α, β, MDE, and sample size
- Checked group balance and experiment validity
- Used z-test to evaluate statistical significance
- Interpreted results and suggested next steps
**Skills:** A/B testing, statistical validation, experiment design
**Tools:** Python, Pandas, SciPy, Statsmodels
**Results:**
- Required sample: 17,441 users per group
- Actual: 15,163 (A), 15,416 (B), test duration: 4 days
- Successful session rate: 30.8% (A) vs 31.8% (B)
- p = 0.00015 → statistically significant, but practically small effect
**Conclusion:** Although the test is statistically valid, the practical gain (1.1%) is minimal. A longer and larger-scale test is recommended.
---
### Data-Delivery — Analytics for a Food Delivery Service
**Code** [`Data-Delivery.sql`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/SQL%20-%20Data%20-%20%D0%B4%D0%BE%D1%81%D1%82%D0%B0%D0%B2%D0%BA%D0%B0.sql)
[`Data-Delivery Dashboard`](https://datalens.yandex/eds6ox6bcyqs0)
**Goal:** Calculate product KPIs, build an interactive dashboard, and provide insights for business development.
**Description:** As an analyst for a food delivery app, I calculated metrics like DAU, conversion, AOV, LTV, and retention using SQL and visualized them in Yandex DataLens.
**Key Contributions:**
- Built SQL queries for KPI calculation
- Created an interactive dashboard with filters and visualizations
- Analyzed seasonal and behavioral trends
- Prepared a business report with recommendations
**Skills:** SQL analysis, metric building, dashboarding, behavioral insights
**Tools:** PostgreSQL (SQL), Yandex DataLens
**Results:**
- DAU showed upward trend in late June
- Conversion improved from 18% to 43%
- Average order value grew by 8.7%
- 7-day retention dropped (5% in May → 3% in June)
- Top LTV restaurants: "Gourmet Delight", "Gastro Storm", "Chocolate Heaven"
**Recommendations:**
- Analyze reasons for low retention
- Launch personalized promotions
- Strengthen partnerships with top-performing restaurants
---
### Property Pulse — Real Estate Market Analysis (St. Petersburg & Region)
**Code:** [`Property Pulse.sql`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/SQL%20-%20Property%20Pulse.sql)
[`Property Pulse Dashboard`](https://datalens.yandex/0hi85722buzen)
**Goal:** Identify promising market segments and visualize key real estate trends in St. Petersburg and Leningrad region.
**Description:** Analyzed listing activity, buyer/seller behavior, and seasonal patterns. Created a dashboard with regional breakdown and activity hotspots.
**Key Contributions:**
- Analyzed listing durations and seasonal peaks
- Identified high-activity zones in the Leningrad region
- Compared buyer vs seller behavior across regions
**Skills:**SQL, seasonal analysis, behavioral trends
**Tools:** PostgreSQL (SQL), Yandex DataLens
**Results:**
- Expensive listings in St. Petersburg remain active longer
- Faster deals in Leningrad region with more affordable options
- Listings peak in autumn; purchases peak in winter and spring
- Top localities: Murino, Kudrovo, Shushary
**Recommendations:**
- Align marketing with seasonal patterns
- Focus on fast-moving regional properties
- Use localized strategies for different market zones
---
### DarkForest — In-Game Purchase Behavior Analysis
**Code:** [`DarkForest.sql`](https://github.com/ElisSorokin/PortfolioProjects/blob/main/SQL%20-%20DarkForest.sql)
**Goal:**Analyze monetization patterns in the fantasy game "Secrets of DarkForest" and identify user segments more likely to spend.
**Description:** Studied player behavior and transaction logs to understand purchasing trends for in-game currency and high-value items.
**Key Contributions:**
- Calculated paying user share by character race
- Analyzed purchase flow for epic items
- Identified behavioral clusters among players
- Prepared a strategic memo for the marketing team
**Skills** SQL, behavioral segmentation, gaming analytics
**Tools:** PostgreSQL (SQL)
**Results:**
- Some races had 1.5–2× higher payment likelihood
- Epic purchases often followed spikes in in-game activity
- Identified 3 user segments: impulse buyers, strategists, regular spenders
**Recommendations:**
- Personalize offers by character race
- Trigger promotions after high activity
- Tailor in-game shop to different player types
---
## Education
University of Rome Tor Vergata:
Bachelor’s Degree in Economics and Finance
Mathematics High School — Specialization in Advanced Mathematics
## Courses:
- Data Analytics – Yandex Practicum, 2025
- Data Engineer: From Zero to Junior – NovaData (via Stepik), 2025
- Managed Service for ClickHouse – Yandex Practicum, 2025
- Docker from Scratch – Karpov.Courses, 2025
- Introduction to Statistics – Institute of Bioinformatics (via Stepik), 2025
## Contacts
- telegram - @elisei_sorokin
- Email - eliseisorokin.work@gmail.com
- LinkedIn: [@elisei sorokin](https://www.linkedin.com/in/elisei-sorokin-8b6106193/)