Back to Projects

E-Commerce Analytics Dashboard

Python · SQL · Neon · Streamlit · Claude Code 2026

Overview

Interactive marketing and customer analytics dashboard backed by a Neon serverless PostgreSQL database. Two-tab Streamlit app: campaign efficiency KPIs (ROAS, CTR, CVR, CPA) with interactive filters, and RFM-based customer segmentation across 8,000+ customers.

Methodology

  • Designed a star schema in Neon PostgreSQL — fact tables for marketing spend and customer RFM scores, dimension tables for campaigns, channels, and customers
  • Queried and aggregated data with SQL via psycopg2; processed results with pandas and numpy
  • Built a two-tab Streamlit dashboard with multi-select filters and caching for performance
  • Visualized campaign ROAS, CTR vs CVR scatter, CPA by channel, and customer segment distribution using Plotly Express
  • Used Claude Code to accelerate schema design, query writing, and dashboard layout

Built With

Python Python
PostgreSQL PostgreSQL
Neon Neon
Streamlit Streamlit
pandas pandas
Plotly Plotly

Screenshots

Campaign efficiency tab — ROAS, CTR/CVR scatter, CPA by channel
Campaign efficiency tab — ROAS, CTR/CVR scatter, CPA by channel
Customer segmentation tab — RFM donut, frequency/value by segment, income mix
Customer segmentation tab — RFM donut, frequency/value by segment, income mix

Results

Identified top-performing campaign types and channels by ROAS and CPA. Segmented customers into Champions, Loyal, Regular, Recent, and At-Risk groups — revealing income bracket composition per segment to guide targeting strategy.

View on GitHub