December 2021 Summaries

Feature stores have become an essential part of ML pipelines for data science teams working with large amounts of data. These feature stores serve as warehouses for documented features that can be used across various ML models, helping organizations save time and resources while ensuring consistency of information and scalability. Automating and centrally managing the data processes powering operational machine learning models, feature stores facilitate quick and reliable development and deployment of features. Some top feature stores available include Tecton, Feast, Hopsworks' Feature Store, and MLflow.