👋 Hi, I'm El khlife Messoud

Data Analyst based in Nouakchott, Mauritania 🇲🇷
Google Certified · Python · SQL · Power BI · Machine Learning

🔗 LinkedIn · Available for Data Analyst / BI roles

---

📊 Portfolio Projects

🛒 Superstore Sales — Customer Segmentation & Profit Modeling

End-to-end retail analytics · PACE framework · Statistical validation

Full analytical pipeline on 9,994 transactions: EDA → T-Test / ANOVA / Chi² → K-Means (k=4) → Ridge regression with K-Fold CV per cluster. Delivered quantified pricing recommendations.

• Segments: VIP (+$5,912), Loyal (+$1,323), Occasional (+$290), At-Risk (−$212)
• Per-cluster R²: 0.57 · 0.99 · 0.32 · 0.80 — all beat global OLS (R²=0.34)
• Business impact: ~$23,318 recoverable profit (Furniture discount cap)
• Stack: Python · scipy · statsmodels · scikit-learn (K-Means, Ridge, K-Fold CV)

→ View project

---

🏦 Bank Customer Churn — XGBoost Predictive Model

🏆 Kaggle Competition entry — Playground Series S4E1

Predicted bank customer churn at competitive level. Trained and benchmarked 3 models (Logistic Regression, XGBoost default, XGBoost optimized).

• Public leaderboard: ~0.887 AUC-ROC ⭐ (surpassing the ~0.88 benchmark)
• Critical segment identified: Germany (37% churn, double France/Spain)
• Method: GridSearchCV · SMOTE on train only · 3-model comparison
• Stack: Python · pandas · scikit-learn · imbalanced-learn · XGBoost

→ View project

---

🎯 Salifort Motors — Employee Retention

Capstone — Google Advanced Data Analytics Certificate

Predicted employee attrition at Salifort Motors (14,999 employees) using Random Forest + SMOTE, following the PACE framework.

• Model: Random Forest with GridSearchCV (5-fold CV)
• Performance: 98% accuracy · ROC-AUC 0.997 · 92% recall on leavers
• Key insight: Workload and tenure drive turnover — NOT salary
• Stack: Python · pandas · scikit-learn · imbalanced-learn · seaborn

→ View project

---

🚴 Cyclistic Bike-Share — Marketing Analysis

Capstone — Google Data Analytics Certificate

Analyzed ride patterns of Cyclistic members vs casual riders to inform marketing strategy converting casual users into annual members.

• Focus: Exploratory data analysis + dashboard design
• Deliverable: Interactive dashboard with actionable marketing recommendations
• Stack: SQL · Data cleaning · EDA · Visualization · Dashboard

→ View project

---

🛠️ Technical Stack

Languages: Python · SQL
Data & BI: pandas · numpy · Power BI · BigQuery · Google Sheets
Machine Learning: scikit-learn · XGBoost · imbalanced-learn · Random Forest · K-Means · Ridge · Linear Regression
Statistics: scipy · statsmodels · hypothesis testing (T-Test, ANOVA, Chi²)
Visualization: matplotlib · seaborn · Power BI
Automation: n8n · Google Workspace APIs
Frameworks: PACE · CRISP-DM

🌍 Languages

French (native) · Arabic (native) · German (C1) · English (B2)

🎓 Certifications

• Google Advanced Data Analytics Professional Certificate
• Google Business Intelligence Professional Certificate
• Google Data Analytics Professional Certificate

---

Projects cover supervised ML (classification · regression), unsupervised ML (clustering), statistical inference, and BI dashboarding — a full Data Analyst toolkit.