featureform/featureform

Health v2 · maintenance only

Adoption not part of the score

1985 stars · 108 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Featureform is a virtual feature store that sits atop your existing data infrastructure and orchestrates it to define, manage, and serve ML model features. It standardizes transformations, features, labels, and training sets with metadata like lineage, variants, and ownership for data science teams.

Use cases

• turn my existing data warehouse into a feature store
• define and serve ML features from notebooks to production
• manage feature definitions and lineage for a data science team
• materialize features on a schedule for online serving
• share and reuse transformations across ML projects
• store and search embeddings for similarity lookups
• build consistent training sets from raw data

When to choose

• you want feature store capabilities without migrating off existing databases and warehouses
• a team needs standardized, versioned feature definitions with ownership and lineage
• you need both offline training sets and low-latency online serving from one source of truth

When to avoid

• you need a fully managed turnkey feature store with no infrastructure to operate
• your ML workflow is small-scale single-user experimentation without collaboration needs
• you only need ad-hoc data transformations without feature serving or metadata management

Facets

framework · maturity active

machine-learning data-science vector-database etl workflow-automation machine-learning data-science python go self-hosted cloud feature-store mlops feature-engineering embeddings data-quality virtual-feature-store data-engineering docker kubernetes

Member repositories

For agents

markdown · JSON · MCP: product_card(name="featureform/featureform")