Portfolio

Data Science & Analytics

One shared framework — churn prediction, segmentation, cohort analysis — proven across five industries with zero code duplication, not five disconnected projects.
education exec findings

Cross-Industry Retention Intelligence Suite

XGBoost • SHAP • KMeans • Streamlit
View Demo
First load may take ~30s to wake up (free-tier hosting)
View GitHub

See It In Action

Education
education exec findings

Segmentation + Retention Trend

27.4% of the population falls in the At-Risk segment — the same shared model, different industry.

Gaming
gaming kpi 3

Segmentation + Retention Trend

Guild members churn at 13.1% vs. 23.2% for non-members — the same model surfaced this without being told to look for it.

Core Capabilities

One Framework, Five Industries

Zero Code Duplication

Music, Education, Healthcare, E-Commerce, and Gaming all run the exact same churn prediction, segmentation, and cohort analysis code — only the data and feature engineering differ.

Explainable AI (SHAP)

Not Just a Score

Every risk score comes with a real explanation of which factors drove it, not a black-box number.

Real-Time Executive Dashboard

Switch Industries Live

KPIs, retention trends, segments, and generated insights recompute instantly when you switch industries in the sidebar.

Technology Stack

Retention Intelligence Suite integrates data engineering, machine learning, and executive dashboarding into one shared framework across five industries.

Data & Analytics

  • Python
  • Pandas
  • NumPy
  • DuckDB

Machine Learning

  • XGBoost
  • scikit-learn
  • SHAP

Dashboard

  • Streamlit
  • Plotly

MLOps & Architecture

  • Plugin-Based Registry
  • YAML-Driven Manifests
  • Shared Framework Design