This project demonstrates an end-to-end Retail Sales Data Analysis using Azure Databricks.
The objective is to analyze retail transaction data to uncover insights related to sales performance, product trends, store efficiency, and customer behavior using Python-based analytics and visualizations.
The project showcases real-world analytics workflows used by Data Analysts and Business Analysts in cloud environments.
Retail organizations generate large volumes of transactional data every day.
Without proper analysis, it becomes difficult to:
- Identify top-selling products
- Track store performance
- Understand customer purchasing behavior
- Optimize pricing and inventory
This project addresses these challenges using Azure Databricks.
- Data Creation / Ingestion
- Data Cleaning & Feature Engineering
- Exploratory Data Analysis (EDA)
- Advanced Data Visualization
- Business Insights Generation
The dataset used in this project is synthetically generated and contains 100 retail transactions.
| Column Name | Description |
|---|---|
| TransactionID | Unique transaction identifier |
| CustomerID | Unique customer identifier |
| Product | Product category |
| Quantity | Number of units sold |
| Price | Price per unit |
| TotalSales | Quantity × Price |
| Store | Store location |
| Date | Transaction date |
| DayOfWeek | Day derived from date |
| Month | Month derived from date |
- Azure Databricks
- Apache Spark (Databricks Runtime)
- Python
- Pandas & NumPy
- Matplotlib & Seaborn
- GitHub
- Created a retail dataset using NumPy
- Converted raw data into Pandas DataFrame
- Derived new features:
TotalSalesDayOfWeekMonth
- Validated data types and integrity
Key EDA operations:
- Product-wise sales analysis
- Store-wise revenue comparison
- Customer spending behavior
- Time-based sales trends
- Distribution analysis of quantity and price
The following charts were created using Matplotlib & Seaborn:
- Bar Charts (Product & Store Sales)
- Line Charts (Daily & Monthly Trends)
- Pie & Donut Charts (Sales Contribution)
- Box & Violin Plots (Price & Quantity Distribution)
- Heatmap (Correlation Analysis)
- Stacked Bar Charts
- Bubble Charts (Price vs Sales vs Quantity)
- Identified top-performing products based on revenue
- Determined highest revenue-generating stores
- Found peak sales periods
- Analyzed customer purchase patterns
- Observed correlation between price, quantity, and sales
This analysis helps retail businesses:
- Optimize inventory management
- Improve store performance
- Identify high-value customers
- Make data-driven pricing decisions
- Enhance marketing strategies
- Upload the dataset to Azure Databricks
- Create a Databricks notebook
- Load data using Pandas or Spark
- Execute analysis and visualization cells
- Interpret insights from charts and metrics
- Real-time data ingestion
- Sales forecasting using ML models
- Customer segmentation
- Integration with Power BI dashboards
- Deployment using Azure Data Factory
Ashwin Kumar
MBA in Data Analytics
Aspiring Data Analyst | Azure Databricks | SQL | Python
This project demonstrates practical implementation of Retail Analytics in Azure Databricks, combining data engineering, analytics, visualization, and business interpretation — making it a strong portfolio project for analytics roles.