Zomato Operations Intelligence & Executive Performance Dashboard

An end-to-end Business Intelligence and Data Engineering project that designs, simulates, structures, and visualizes operations logistics, financial performance, and customer cohorts for Zomato's Operations Leadership Team.

---

Project Architecture

This project is built using a classic Star Schema data warehouse architecture. A custom data generation pipeline simulates 18 months of historical operations records, which are then cleaned, modeled, and visualized inside Power BI Desktop:

[python/config/settings.py] ──> [python/business_rules/rules.py]
                                       │
                                       ▼
                     [python/generators/run_pipeline.py]
                                       │
                                       ▼
                           [data/synthetic/*.csv]
                                       │
                                       ▼
                            [Power Query ETL (M)]
                                       │
                                       ▼
                            [Power BI Data Model]
                                       │
                                       ▼
                       [Interactive 3-Page Dashboard]

---

Repository Structure

• 📁 data/synthetic/ — Generated relational CSV files (Cities, Customers, Restaurants, Riders, Promotions, Orders, Payments).
• 📁 python/ — Synthetic data generator scripts and custom business simulation rules (seasonality, growth rates, prep times, SLAs).
• 📁 sql/schema/ — Production DDL schema scripts for MySQL compatibility.
• 📁 powerbi/ — Tabular data model report (.pbix), custom dark flat theme (zomato_theme.json), and graphical layout assets.
• 📁 documentation/ — Multi-page documentation covering Business Requirements, Data Dictionaries, DAX Formulas, and setup guides.

---

Dataset Scope (18 Months)

• Temporal Coverage: January 1, 2024 – June 30, 2025.
• Cities: 15 Indian cities classified into Tier 1, Tier 2, and Tier 3 markets.
• Orders Fact Table: 119,994 transactional records.
• Payments Fact Table: 123,275 transaction attempts (incorporates digital failures and Cash on Delivery).
• Active Restaurant Partners: 1,287 online partners grouped by tier (Gold, Silver, Bronze).
• Active Rider Fleet: 669 delivery partners using Motorcycles, Scooters, and Bicycles.

---

Key Performance Indicators (KPIs) Modeled

The project maps and calculates 38 DAX metrics organized across three consolidated views:

1. Executive & Financial Intelligence

• Total Revenue / GMV: Tracking financial growth and platform margins.
• Average Order Value (AOV): Monitoring average cart spends.
• Promotion ROI: Quantifying discount campaign profitability.
• Payment Success Rate: Gateway performance auditing.

2. Operations & Delivery Analytics

• On-Time Delivery Rate (OTDR): SLA tracking (90% target) showing rainy-season monsoon delays.
• Average Prep Time: Kitchen latency audits across restaurant tiers.
• Orders Heatmap: Grid mapping time-of-day order density (lunch/dinner rushes).
• Rider Utilization: Deliveries completed per active rider per day.

3. Customer Behavior & Demographics

• Customer Cohort Segments: Profiles users into New, Active, Lapsed, Churned.
• Customer Churn Rate: Monitoring user attrition metrics.
• CSAT Score: Customer rating distributions (1-5 stars).
• Demographic Segmentation: Slice orders by customer age groups and gender.

---

Setup & Data Generation

To regenerate the synthetic dataset or modify the business rules:

1. Initialize Python virtual environment:

   python -m venv venv
   .\venv\Scripts\activate

2. Install dependencies:

   pip install -r requirements.txt

3. Run the generation pipeline:

   python python/generators/run_pipeline.py