A library focused on creating data pipelines with an emphasis on composability, portability, and performance, integrating various data processing and machine learning tasks.

A feature that allows scoring XGBoost models using a Rust-based inference engine directly within DataFrames.