https://github.com/devanshsahu47/customer-sales-dashboard
Interactive Tableau dashboard analyzing customer sales trends by age, gender, region, product category, and discount behavior. Built using real sales data to derive actionable business insights.
https://github.com/devanshsahu47/customer-sales-dashboard
Last synced: 5 months ago
JSON representation
Interactive Tableau dashboard analyzing customer sales trends by age, gender, region, product category, and discount behavior. Built using real sales data to derive actionable business insights.
- Host: GitHub
- URL: https://github.com/devanshsahu47/customer-sales-dashboard
- Owner: devanshsahu47
- Created: 2025-06-22T06:32:19.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-06-22T14:15:17.000Z (about 1 year ago)
- Last Synced: 2025-07-26T13:44:25.544Z (11 months ago)
- Size: 25.3 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🤖 Customer Sales Dashboard (AI-Augmented)
This project presents an interactive Tableau dashboard, developed using **GenAI-powered EDA and prompt engineering**, to analyse customer behaviour, sales distribution, and key performance trends. By combining traditional BI tools with **AI assistants like PowerDrill**, the project showcases how automated analysis can enhance business decision-making.
---
## 📌 Objective
To generate fast, reliable, and explainable sales insights using GenAI tools and visual storytelling. This includes:
- AI-powered exploration of large transactional datasets
- Prompt-based segmentation and pattern discovery
- Interactive Tableau visualisation for key business questions
---
## ⚙️ AI-Powered Workflow
- 🔍 **EDA with PowerDrill (GenAI Tool)**: Used natural language prompts to generate summaries, correlation patterns, and detect outliers or segment-specific trends.
- 💬 **Prompt Engineering**: Crafted structured prompts to auto-generate analysis on customer segments, discount impact, and category-wise performance.
- 📊 **Visualisation**: Insights from AI were translated into human-readable dashboards via Tableau for stakeholder-ready storytelling.
---
## 📈 Dashboard Insights
### 🔹 Month-Wise Revenue
- Revenue peaked in **October 2020 ($57.7M)** and **April 2021 ($36.7M)**.
- Dips in **February 2021 ($4.4M)** and **July 2021 ($19.0M)** suggest seasonal or operational shifts.
### 🔹 Revenue by State
- Top states:
**Texas ($15.5M)**, **California ($13.9M)**, **Florida ($11.4M)**
### 🔹 Age-Wise Sales
- Dominant segments:
- **30–40** age group – $44.8M
- **60–70** – $41.1M
- **20–30** – $39.9M
### 🔹 Regional Revenue Distribution
- **South** – 38.37%
- **Midwest** – 26.93%
- **West** – 17.60%
- **Northeast** – 17.10%
### 🔹 Gender-Category Behaviour
- **Males** prefer: Computing, Men’s Fashion, Entertainment
- **Females** prefer: Women’s Fashion, Others, Entertainment
### 🔹 Quantity vs Discount Trend
- Most orders occur at **<20% discount**.
- Some high-quantity orders at higher discounts indicate **price sensitivity** in selective segments.
---
## 📦 Dataset
`sales_06_FY2020-21.csv` – Contains customer purchase records with:
- Order Date, State, Region
- Age, Gender, Category
- Quantity, Revenue, Discount %
---
## 🧠 Tools & Technologies
| Tool | Purpose |
|----------------|----------------------------------|
| PowerDrill | GenAI-driven EDA and summarisation |
| Tableau | Visual storytelling & dashboarding |
| Excel/Sheets | Light data cleanup |
| GitHub | Version control and publishing |
---
## 🔍 Key Business Takeaways
- **Texas and California** are highest-performing states.
- Customers aged **30–70** drive most revenue.
- **Entertainment and Fashion** dominate sales categories.
- **South region** outperforms others in total revenue.
- Discounts don't significantly drive volume, except in certain outlier cases.
---
## 📌 AI + BI Value Add
This project demonstrates how **GenAI tools can accelerate and scale data analysis**. With minimal coding, high-level insights were derived using structured prompting, allowing the analyst to focus more on decision-making than manual query writing.
---
## 📝 License
Open-source under the **MIT License**.
---
