Aleksandr Gordeev
Data Analyst · Business Analyst
Turning data into real business insights for making better decisions
Analysed a lending portfolio to identify growth opportunities and risk control improvements, delivering an interactive Power BI dashboard for strategic decision-making.
Dashboard overview
Customer segmentation
Profit analysis
Risk analysis
- Identified most profitable states and customer groups by income band and loan purpose — wedding loans showed a 17.6% higher profit margin than the overall portfolio.
- Uncovered grade-C loans as the best risk-adjusted return opportunity — 7.11% return while representing only 27.9% of the portfolio, highlighting scalable growth potential.
- Deployed interactive dashboard integrating Portfolio at Risk, Loan YoY growth, weighted interest rate, and profitability KPIs for data-driven risk and lending strategy.
End-to-end NLP and machine learning pipeline analysing British Airways customer reviews and predicting booking completion to support marketing and customer experience strategy.
Sentiment table
Sentiment distribution
Predictive model — feature importance
- Scraped and processed third-party review data with BeautifulSoup & NLTK, mapping satisfaction distribution across positive, neutral, and negative segments using VADER sentiment.
- Built a Random Forest classifier achieving 85.44% accuracy in predicting booking completion — identifying destination, purchase lead, and flight length as key behavioural drivers.
- Delivered visualised insights structured for non-technical stakeholder communication.
Analysed a large credit card transaction dataset of 1.85 million records including transaction timestamps, amounts, merchant details, customer demographics, and geospatial information to surface fraud risk areas and customer behaviour trends.
Segmentation by age
Spending by age
Segmentation by purchase category
Average spend by category
Total spend by job title
Avg spend by job role
Transaction volume by US state
Fraud transaction volume by location
Top fraud merchants
Top fraud customers
- Identified business growth opportunities through complex customer segmentation by age group, job role, and purchase category — revealing key spending behaviour patterns.
- Mapped transaction geographic distribution across US states, identifying high-risk fraud regions and the merchants and customers with the highest fraudulent transaction counts.