Feast Feature Store

Open-source feature store for operationalizing machine learning features for training and serving

MLOpsFree

Feast (Feature Store) is an open-source feature store that helps ML teams operationalize features for training and serving machine learning models. It provides a central registry for defining and documenting features, point-in-time correct data retrieval for training dataset generation, and low-latency feature serving for production model inference. Feast connects to existing data infrastructure—Snowflake, BigQuery, Redis, DynamoDB—rather than requiring data migration. ML platform teams at companies building production ML systems use Feast to solve the training-serving skew problem, ensuring models receive identical features during training and production serving. As an open-source project with a large community, Feast has become the de facto standard open-source feature store.

Key Features

• ✓Feature registry
• ✓Point-in-time retrieval
• ✓Low-latency serving
• ✓Multiple data sources
• ✓Open-source