Vendor Performance Analysis

Overview

The Vendor Performance Analysis project is an end-to-end data analytics solution designed to evaluate vendor performance using procurement, inventory, sales, and pricing data. The project ingests raw CSV datasets into a MySQL-backed relational database, performs aggregations and KPI calculations, and produces analytical outputs suitable for business decision-making and executive dashboards.

This repository intentionally contains only source code, notebooks, and configuration files. All large datasets, database files, and output artifacts are hosted externally and referenced below.

---

Business Objectives

• Consolidate vendor-related data from multiple operational sources
• Evaluate vendor performance using financial and operational KPIs
• Identify high-performing and under-performing vendors and brands
• Enable downstream visualization using Power BI connected directly to MySQL

---

Power BI Dashboard Preview

Primary Data Source: MySQL
The Power BI dashboard connects directly to MySQL tables generated by the Python ingestion and transformation pipeline.

---

Project Architecture

Data Flow

1. Raw CSV files -> data/ directory (downloaded externally)
2. Ingestion into SQLite/MySQL using Python
3. Vendor-level aggregation and KPI computation
4. Final summary tables stored in MySQL
5. Visualization using Power BI

---

Repository Structure

vendor-performance-analysis/
│
├── data/                       # Ignored (download separately)
├── logs/                       # Application logs
│   └── ingestion_db.log
│
├── eda.ipynb                   # Exploratory Data Analysis
├── ingestion_db.ipynb          # Notebook-based ingestion
├── ingestion_db.py             # Raw data ingestion script
├── vendor_summary.py           # Vendor KPI aggregation logic
├── performance_analysis.ipynb  # Performance analysis
│
├── vendor_analysis.db          # Ignored (SQLite – dev only)
├── vendor_sales_summary.csv    # Ignored (final output)
├── vendor_analysis_dashboard.pbix  # Ignored (Power BI file)
│
├── .gitignore
├── requirements.txt
└── README.md

---

External Data & Outputs

All datasets and large artifacts are hosted on Google Drive.

Google Drive Link: https://drive.google.com/drive/folders/11wFFbi7JJ3dBptVhIKTwmVTtT1_B0Aay

Drive Folder Contents

• Database db file – SQLite database generated via Python

• Dataset – All raw CSV files (place inside data/)

• Sales summary csv – Final aggregated output csv

• Power BI files – Dashboard .pbix file and screenshots

---

Database Tables

The ingestion pipeline programmatically creates the following tables:

• sales

• begin_inventory

• end_inventory

• purchase_prices

• purchases

• vendor_invoice

• vendor_sales_summary

---

Key Metrics Computed

• Gross Profit

• Profit Margin (%)

• Stock Turnover

• Sales-to-Purchase Ratio

• Freight Cost Impact

---

Logging

• All ingestion and processing steps are logged

• Logs are stored in the logs/ directory

---

Skills & Technologies Used

This project demonstrates:

• ETL & Data Engineering: Python, Pandas, SQLAlchemy, MySQL
• Database Design: Relational schema, table creation, data aggregation
• Data Analysis & Visualization: Jupyter Notebooks, KPI computation
• Business Intelligence: Power BI dashboard creation and reporting
• Version Control: Git & GitHub project management
• Logging & Debugging: Structured logging of ingestion and transformation

---

Future Enhancements

• Parameterized database configuration via .env

• Automated MySQL deployment

• Incremental ingestion

• Data validation checks

• CI/CD integration