Customer Intelligence & Revenue Optimization
A full-stack customer intelligence dashboard built on the Olist Brazilian e-commerce dataset (2016-2018), analyzing 96,096 unique customers and $15.42M in delivered revenue. Every number on the dashboard is computed live — a FastAPI backend runs raw SQL against MySQL on each request, and a Next.js + Recharts frontend renders the result. The dashboard answers four real business questions: who the best customers are (RFM segmentation), whether customers come back (repeat-purchase & cohort retention), what products get bought together (product affinity), and where revenue is leaking (canceled/unavailable order analysis) — each module ending in a data-backed recommendation.
Model Accuracy
N/A
Model Type
SQL Analytics Engine
Category
SQL & Analytics
Tech Stack
6 Tools
Key Features
- ✓RFM segmentation scoring 93,357 customers into Champions, Loyal, New, At Risk, Need Attention & Lost
- ✓Repeat-purchase & cohort retention analysis across all delivered customers
- ✓Product-affinity mining of category pairs frequently bought together
- ✓Revenue-leakage detection tracing canceled/unavailable orders back to product category
- ✓Live FastAPI + SQLAlchemy backend — every metric queried from MySQL on request
- ✓Recharts-powered Next.js dashboard with a recommendation card per module
Technologies
Project Screenshots
Performance Metrics
Total Customers
96,096
Revenue Analyzed
$15.42M
Avg Order Value
$159.86
Revenue at Risk
$270K