Overview
A 3,900-record analytics case study that turns customer shopping data into answers for 13 business questions, then carries those answers into a Power BI reporting experience.
Evidence
Python / Pandas prepared the dataset; PostgreSQL hosted the cleaned data; SQL answered the analytical questions; Power BI presented the reporting layer.
Limitations
This portfolio demonstration uses static project results. It does not connect to a live database or expose confidential customer data.
Analysis overview
