Data Science & Analytics Projects
Comprehensive portfolio showcasing expertise in machine learning, business intelligence, statistical analysis, and data-driven strategic insights
Interactive Market Share Dashboard
Comprehensive Power BI dashboard analyzing FMCG market dynamics across multiple product categories, regions, and time periods. Provides real-time competitive intelligence and market performance tracking.
Customer Churn Prediction Model
Advanced machine learning model predicting customer churn in telecommunications. Achieved 69.6% ROC-AUC using Gradient Boosting, enabling proactive retention strategies and customer lifetime value optimization.
Social Media Virality Impact Analysis
Strategic analysis of TikTok and viral video marketing effectiveness across product categories. Identifies optimal platform-content combinations and quantifies ROI for viral marketing campaigns.
E-commerce A/B Testing & Conversion Optimization
Designed and analyzed 15+ A/B tests for an e-commerce platform, optimizing checkout flow, pricing strategies, and UI elements. Implemented Bayesian A/B testing framework achieving 23% increase in conversion rate.
Multi-Language Sentiment Analysis & Topic Modeling
NLP system analyzing 500K+ customer reviews across 12 languages using fine-tuned BERT models. Achieved 91% sentiment accuracy and identified 5 critical product issues, leading to 18-point NPS improvement.
Customer Lifetime Value Prediction & Segmentation
Probabilistic CLV model for 250K+ customers using BG/NBD and Gamma-Gamma models. Enabled targeted retention strategies reducing churn by 28% in high-value segments and achieving 3.4x marketing ROI.
Hybrid Recommendation System with Deep Learning
Production-scale recommendation engine serving 2M+ users using two-tower neural network architecture. Achieved 34% CTR increase with sub-50ms inference latency through hybrid collaborative filtering approach.
Multi-Horizon Demand Forecasting System
Ensemble forecasting system for 5,000+ SKUs combining Prophet, LSTM, and XGBoost. Achieved 87% accuracy at 4-week horizon, reducing inventory costs by €3.1M annually while improving stock availability.
Real-Time Fraud Detection with Graph Neural Networks
GNN-powered fraud detection system processing 1M+ daily transactions. Achieved 94% detection rate with 0.8% false positive rate, preventing €8.7M in fraudulent transactions annually.
Real-Time Anomaly Detection for IoT Sensors
Production MLOps system processing 10M+ daily sensor readings using Isolation Forest and LSTM autoencoders. Achieved 67% false positive reduction with 48-hour advance failure warnings.
About These Projects
These projects demonstrate comprehensive data science and analytics capabilities, combining advanced technical skills with strategic business thinking. Each project showcases:
- →Technical Expertise: Machine Learning, Deep Learning, NLP, Time Series, Statistical Analysis
- →Business Impact: Quantified results, ROI analysis, strategic recommendations
- →Production Skills: Scalable systems, real-time processing, MLOps practices
- →Communication: Clear visualizations, actionable insights, stakeholder presentations
All projects include rigorous methodology, professional visualizations, and measurable business outcomes.
Interested in Working Together?
Let's discuss how data-driven insights can drive your business forward.
Get In Touch