Kroger Dashboard Web Application

This repository contains the source code for a data analytics web application developed as part of a group project assignment. The project leverages modern data engineering practices and cloud technologies (Azure) to provide interactive insights into retail customer behavior based on anonnymized 84.51°/Kroger data.

Project Overview

The goal of the project is to explore real-world retail data and answer key questions regarding customer engagement and demographic factors. Specifically, the application helps answer:

• How does customer engagement change over time? (e.g., changes in spend, basket items, and overall engagement trends)
• Which demographic factors (e.g., household size, presence of children, income) appear to affect customer engagement?

Using sample household transactions and demographic data, the application process and aggregates data with Polars and visualizes the results using interactive Plotly charts. The application also provides an interactive data table (powered by grid.js) to browse through transaction records with features like search, sorting, and pagination.

Features

• Interactive Dashboard:
  • Displays bar and pie charts illustrating total and average spending segmented by Age Range, Marital Status, Income Range, and Number of Children.
• Dynamic Data Table:
  • Display of transaction details with search, sorting and pagination.
• User Authentication:
  • Simple registration and login pages that allow users to create accounts and securely access the dashboard.
• Data Processing:
  • Fast data aggregation and manipulation using Polars.
• Conditional Azure Integration:
  • The code includes Azure Blob Storage and PostgreSQL integration for cloud deployment For local development, this is disabled by default (controlled by the USE_AZURE environment variable).

Demo

Below is a short demo (<1 minute) showcasing:

• Registration and account creation.
• Dashboard functionality (charts and interactive table with search, sorting, and pagination). Login with newly registered credentials.

Demo

Setup and Installation

Follow these steps to run the application locally:

1. Clone the Repository:

git clone https://github.com/yourusername/kroger-dasboard.git
cd kroger-dasboard

2. Create a Virtual Environment & Install Dependencies:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt

3. Configure Environment Variables: Create a `.env file in the project root with the following contents (adjust values as needed):

# PostgreSQL settings (if using for user management)
PG_DATABASE=dasboard
PG_USER=myuser
PG_PASSWORD=mypassword
PG_HOST=localhost
PG_PORT=5432

# Application port
PORT=5000

# Toggle Azure integration (set to "false" for local demo)
USE_AZURE=false

4. Run the Application:

python app.py

By default the application will run in debug mode on http://0.0.0.0:5000/

Usage

Registration & Logic:

• On first visit, you'll be directed to the registration page. Create an account using your email, username, and password.
• After registration, use your credentials on the login page to access the dashboard.

Dashboard:

• View interactive charts showing spending and engagement metrics segmented by demographics.
• Browse the transaction data via the dynamic table. Use the search box to filter results by household number, and sort columns as needed

Data Processing:

• The backend aggregates and processes data from detailed-data.csv using Polars, providing real-time charts and statistics.
• Although the repository contains sample login data stored in CSV, the application includes code to connect to a PostgreSQL database for user management, which can be enabled with proper configuration.

Local vs. Cloud Deployment

For local development and demo purposes, Azure integration is disabled by default via the USE_AZURE environment variable. The app uses local CSV files (e.g., data/detailed-data.csv and data/logins.csv) for data processing and PostgreSQL for user management.

To enable Azure integration for a cloud deployment:

• Set USE_AZURE=true in your .env file.
• Provide the necessary Azure credentials (e.g., ACCOUNT_NAME, ACCOUNT_KEY, CONTAINER_NAME, AZURE_BLOB) in the environment.

Note: Additional configuration may be required when deploying to an Azure environment.

Additional Context

The project was developed based on the following assignment brief:

Project Abstract
Organizations today need to leverage modern data engineering as data volume and complexity grow. In retail, machine learning can analyze extensive datasets to forecast product demand and optimize pricing strategies. This project explores real-world retail data from 84.51°/Kroger to build an interactive web application on Azure, making the customer's shopping experience easier.

Key Questions:
  1. How does customer engagement change over time?
  2. Which demographic factors affect customer engagement?

This README summarizes the core components of the assignment and provides insight into the structure and functionality of the web application.