Feat is a program for ingesting financial time series data (e.g., trades from an exchange) and using that data to generate samples and features for research, backtesting, and developing machine learning models. It is written in Rust and optimized for performance, enabling researchers to move faster on experimenting with new ideas, or simply to generate input data for existing models more quickly.
It is inspired by the work of Lopez de Prado and mlfinlab. Feat is still early days and needs your help to shape its direction! Make sure to check out Future and give us feedback about what you would like to see developed.