Code
End-to-End Retail Sales Analytics using Microsoft Fabric

End-to-End Retail Sales Analytics using Microsoft Fabric

Microsoft Fabric

Retail Sales Analytics (Microsoft Fabric + Power BI + SQL)

Overview

Built an end-to-end Retail Analytics solution using Microsoft Fabric, SQL, and Power BI to convert raw sales data into actionable business insights. The project focuses on improving visibility into sales, profitability, shipping efficiency, and customer behavior.

Key Highlights

  • Designed a modern data pipeline (Bronze → Silver → Gold)
  • Implemented star schema for scalable analytics
  • Built Direct Lake semantic model (no data duplication, high performance)
  • Developed interactive Power BI dashboards

Tech Stack

  • Microsoft Fabric (Lakehouse, Dataflow, Direct Lake)
  • SQL (analysis via SQL endpoint)
  • Power BI + DAX
  • Python (Pandas) for data transformation

Pipeline Summary

  • Bronze: Raw CSV ingestion into Lakehouse
  • Silver: Data cleaning, type conversion, feature engineering using Pandas
  • Gold: Star schema with:
    • fact_sales
    • dim_customer, dim_product, dim_location, dim_shipping, dim_date
  • Analytics: 15+ SQL queries (CTEs, window functions, joins)
  • Semantic Layer: Direct Lake model
  • Visualization: Power BI dashboards

Key Metrics

  • Total Sales
  • Total Profit
  • Profit Margin %
  • Average Order Value
  • On-Time Delivery %

Business Insights

  • High sales ≠ high profit → margin optimization needed
  • Loss-making categories identified → pricing strategy improvement
  • Regional shipping delays → logistics optimization opportunity
  • Top customers drive majority revenue → retention focus

Architecture

End-to-End Retail Sales Analytics using Microsoft Fabric Architecture

Dashboard

  • Executive Overview
  • Profitability Analysis
  • Shipping Performance
  • Customer Insights

📌 Full Project : Click Here