This project is part of Microsoft’s Data Science certification module and focuses on uncovering insights from sales, retail, and marketing datasets using Python. Built with a strong analytical backbone and branded visual storytelling, the project demonstrates predictive modeling, customer segmentation, and campaign impact analysis.
sales-retail-analysis/ ├── data/ # Raw and cleaned datasets ├── notebooks/ # Jupyter walkthroughs ├── visuals/ # Branded dashboards and plots ├── models/ # Saved ML models (.keras / .tflite) ├── scripts/ # Reusable Python scripts ├── README.md # Project overview └── requirements.txt # Package dependencies
Forecast future sales with regression models
Identify top-performing categories and regions
Explore inventory optimization
Forecast peak shopping periods using time-series methods
Perform clustering and sentiment analysis
Visual storytelling with branded dashboards
pandas, NumPy for preprocessing
matplotlib, seaborn for visualization
scikit-learn for modeling and clustering
.keras and .tflite formats for lightweight model deployment
Strategic insights for retail and marketing teams
Reusable templates for data-driven storytelling
Lightweight deployable models for demo purposes
Built by Leonard Phokane, combining data science, ethical AI, and visual branding to empower real-world impact in local communities.
Beyond this certification project, I’ve built a range of community-driven and ethically grounded tech solutions. Explore my portfolio to see how I merge creative coding, AI, and blockchain with social impact:
🌍 Portfolio Website → phokane-creative-code.lovable.app
KaziLink: A job-matching platform uplifting local communities
COMPAS Audit: An ethical AI tool promoting fairness and transparency
Recycle Dataset: ML-powered waste classification for environmental impact
📱 Skills in Flutter, TensorFlow Lite, Blockchain, Firebase, and more
“Technology is most powerful when it amplifies human potential and creates opportunities for those who need them most.” — Leonard Phokane