Customer Segmentation Clustering Model

I’m preparing a machine-learning workflow focused on customer segmentation and need a robust clustering model built in Python. The raw data are already available and combine transactional history with basic demographic fields; I’ll share a sample as soon as the project starts. Your job is to explore the data, select and justify an appropriate clustering approach (K-Means, hierarchical, DBSCAN—whichever proves most meaningful), and deliver a clear explanation of each segment’s defining traits so I can target them in future marketing campaigns.

You’re free to leverage scikit-learn, pandas, NumPy, and matplotlib/Seaborn for exploration and visualisation, and Flask or Streamlit if you see value in exposing the results through a lightweight dashboard. Please optimise hyper-parameters, compare at least two algorithms, and support your final choice with metrics such as silhouette score or Davies-Bouldin.

Deliverables
• Reproducible Jupyter notebook (or .py script) showing data prep, modelling, evaluation and visualisations
• A concise report (PDF or Markdown) summarising methodology, metrics, and practical insights for each segment
• Exportable cluster labels (CSV) mapped to the original customer IDs
• Optional: lightweight interactive dashboard for quick, non-technical review of segment profiles

I’m looking for clean, well-documented code and actionable insight rather than just numbers, so include comments and brief explanations directly in the notebook. If you’ve tackled similar segmentation work before, a quick note or link to past results will help me understand your approach.