Skip to content
Latchkey

Feature Store - CI/CD Glossary Definition

A feature store centrally computes and serves ML features so training and inference use identical, versioned feature values.

A feature store is a system that computes, stores, and serves machine-learning features consistently for both training and inference, preventing training-serving skew.

Feature stores solve a core MLOps problem: making sure the features a model trains on match the features it sees in production.

Online vs offline

The offline store serves historical features for training; the online store serves low-latency features for real-time inference. Both derive from the same definitions.

Related guides

Run this faster and cheaper on Latchkey managed runners. Start free →