« Back to Glossary Index

Feature store is a data system used in Machine Learning (ML), used as a centralised hub for storing and organising commonly used features, which can be accessed by several different ML models (or different versions of the same ML model). Having all features in a central location makes data pre-processing and feature engineering into a streamlined and coherent process, and makes features available for reuse in the development of future (versions of) machine learning models, avoiding repeating their computation.