Skip to content

E-commerce Dashboard

Python Pandas SQLite Jupyter Chart.js

Overview

Final project for the Analytics Deep Dive module (Profissão Ciência de Dados). Consolidates e-commerce transaction and customer data via SQL and displays the results in an interactive HTML dashboard with dynamic filters.

Highlights

  • Complete SQL pipeline: ingestion, normalization, INNER JOIN and export
  • Self-contained HTML dashboard — 296 records embedded, no server required
  • 7 metrics and interactive visualizations with filters by gender, state and category
  • Raw data normalized (European format 72,93 → decimal 72.93)

Project Structure

ecommerce-dashboard/
├── data/
│   ├── raw/                 # Source data (transactions and customers)
│   └── processed/           # Consolidated dataset (JOIN result)
├── notebooks/
│   └── Profissao Cientista de Dados M26 Projeto.ipynb
├── src/
│   └── dashboard.html       # Interactive dashboard (Chart.js)
├── requirements.txt
└── README.md

Tech Stack

Technology Version Role
Python 3.13 Main language
pandas latest Data manipulation and export
SQLite (sqlite3) built-in In-memory database for the JOIN
Jupyter Notebook latest Interactive data pipeline
Chart.js 4.4.0 Dashboard visualizations

Data Pipeline

TB_TRANSACOES_PROJETO_ECOMM.csv   TB_CLIENTES_PROJETO_ECOMM.csv
              │                                │
              ▼                                ▼
     ┌────────────────┐             ┌────────────────────┐
     │  Raw ingestion │             │   Raw ingestion    │
     │  Normalization │             │                    │
     │  Price: ,→.    │             │                    │
     └────────┬───────┘             └─────────┬──────────┘
              │                               │
              ▼                               ▼
     ┌─────────────────────────────────────────────────┐
     │              In-memory SQLite                   │
     │  INNER JOIN on id_client                        │
     └─────────────────────────┬───────────────────────┘
                               │
                               ▼
              dados_ecommerce_final.csv (296 records)
                               │
                               ▼
                      dashboard.html (self-contained)

JOIN Rationale

INNER JOIN was chosen because the goal is to analyze transactions linked to identified customers. Transactions without a matching customer record (orphaned rows) add no value to demographic or geographic metrics — resulting in 296 consolidated records.

Dashboard

The src/dashboard.html file is self-contained: all 296 records are embedded directly in the HTML, with no server or live database connection required.

Displayed metrics:

Metric Type
Total sales and average ticket KPI cards
Transactions and unique customers KPI cards
Geographic distribution by state Bar chart
Demographic profile by gender Doughnut chart
Revenue by product category Horizontal bars
Price distribution Histogram (10 bins)
Top roles by transaction count Horizontal bars

Interactive filters: gender, state and category — all KPIs and charts update when a filter is selected.

How to Run

Data pipeline (optional)

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Open the notebook and run all cells. The file data/processed/dados_ecommerce_final.csv will be generated.

Dashboard

Option A — local server (recommended):

python3 -m http.server 8000

Open http://localhost:8000/src/dashboard.html.

Option B — open directly in browser:

Double-click src/dashboard.html or use File > Open.