I solve real-world business challenges with data tools & techniques

Dift

Python, open source

Dift is an open-source CLI tool that helps data professionals compare two datasets and instantly understand what changed, why it matters and whether the new data is safe to trust.

Dift is open to contributions.

Credit Scoring & Loan Decision System

Python, Streamlit, Cloudflare R2, Docker

This is an End-to-end ML system predicting credit default risk to support loan decisions. Built Logistic Regression and XGBoost models with feature engineering and missing value handling. Developed interactive Streamlit dashboard with SHAP-based explainability and actionable business insights.

SegmenAI - Customer Segementation App

Python, scikit-learn, cloudpickle, Streamlit, Docker, MinIO, Cloudflare R2

This is an end-to-end customer segmentation pipeline using RFM and KMeans to help businesses identify high-value, occasional, and inactive customers, with a Streamlit dashboard for interactive visualizations and personalized recommendations.

Credit Card Fraud Detection App

Python, scikit-learn, joblib, Streamlit

This is an end-to-end ML system to detect fraudulent credit card transactions on an imbalanced dataset, experimenting with Logistic Regression, Random Forest, and XGBoost, and deployed an interactive Streamlit dashboard with SHAP explainability for real-time monitoring and risk reduction.

Data Quality Assessment App

Python & Streamlit

This project is an interactive Streamlit dashboard that analyzes and visualizes data quality metrics and error clusters in transactional datasets.