How nagendraseelam/retail-customer-intelligence works
The repository is a retail customer-intelligence project: a notebook documents transaction analysis, RFM segmentation, and high-value prediction, while a Streamlit app accepts customer metrics and returns a prediction with probability. The README documents the notebook’s analysis and modeling stages; only the app source was sampled, so notebook-stage connections rely on README statements rather than inspected implementation.
Main entry points: Retail_Customer_Intelligence.ipynb Data Cleaning · app.py Streamlit Input · xgboost_high_value_model.pkl XGBoost Model · features.pkl Feature Names
Components 12
Business Analysis
- Data Cleaning
Retail_Customer_Intelligence.ipynb - Business KPIs and EDA
Retail_Customer_Intelligence.ipynb
Customer Segmentation
- RFM Analysis
Retail_Customer_Intelligence.ipynb - K-Means Segments
Retail_Customer_Intelligence.ipynb
Value Prediction
- Customer Features
Retail_Customer_Intelligence.ipynb - Model Training and Evaluation
Retail_Customer_Intelligence.ipynb - XGBoost Model
xgboost_high_value_model.pkl - Feature Names
features.pkl
Prediction App
- Streamlit Input
app.py - Prediction and Probability
app.py
Other components
- Online Retail Data
- Business User
Connections 12
- Online Retail Data → Data Cleaning: provides transactions
README.md - Data Cleaning → Business KPIs and EDA: prepares data for
README.md - Data Cleaning → RFM Analysis: prepares data for
README.md - RFM Analysis → K-Means Segments: clusters customers
README.md - Data Cleaning → Customer Features: supports feature creation
README.md - Customer Features → Model Training and Evaluation: trains and evaluates
README.md - Model Training and Evaluation → XGBoost Model: serializes model
README.md - Customer Features → Feature Names: provides feature names
- Business User → Streamlit Input: enters customer metrics
README.md - Streamlit Input → XGBoost Model: loads model
app.py - Streamlit Input → Feature Names: loads feature names
app.py - Streamlit Input → Prediction and Probability: predicts and displays
app.py