A production-grade financial risk analysis system built with Deepbox, demonstrating comprehensive use of statistics, linear algebra, and DataFrame operations.

A neural image classification project built with Deepbox, demonstrating an end-to-end MLP workflow for the digits dataset.

A production-grade customer churn prediction system demonstrating classical machine learning with Deepbox.

A time series forecasting project for stock prices using synthetic market data, feature engineering, and regression baselines.

A collaborative filtering recommendation system using user-item ratings, clustering, and dimensionality reduction.

A text classification system for sentiment analysis demonstrating NLP preprocessing and classification.

A production-style transaction monitoring workflow built around the biggest new v1.0.0 additions: calibrated classification, anomaly detection, feature inspection, and operations-ready reporting.

A production-style text operations pipeline for support routing, using the v1.0.0 text preprocessing stack, model selection helpers, and reporting outputs.

A production-style experimentation and rollout analysis workflow built around the v1.0.0 inference surface: confidence intervals, bootstrap uplift analysis, multiple-comparison correction, KDE diagnostics, and power planning.