Freedom251 › Feature Store Software
Databricks Feature Store
What Databricks Feature Store is
Databricks Feature Store is a cloud registry for features used in AI and machine-learning models. Registering features and models in Unity Catalog supports governance, lineage, point-in-time joins, and sharing and discovery across workspaces. Features can be authored as declarative Feature Views managed through Databricks pipelines or as feature tables populated in Delta tables; Feature Views and their materialization capability are in Public Preview. Models trained with these features track lineage and look up the latest values at inference. Databricks says using the same feature computations for training and inference eliminates training/serving skew. Serving endpoints and online stores support real-time applications, and the overview describes millisecond latency. Third-party online store integrations include Amazon DynamoDB, Amazon Aurora (MySQL-compatible), and Amazon RDS MySQL, with availability varying by mode. The paid service has a free trial and monthly usage-based billing. It requires a workspace enabled for Unity Catalog, and its APIs cannot be called from local or other non-Databricks environments, although local IDE development and unit testing are supported.
Who it suits
Databricks Feature Store suits teams managing model features within a Unity Catalog-enabled Databricks workspace. It may fit teams that need feature lineage, point-in-time joins, or online serving for real-time applications.
What it does well
- Supports governance, lineage, and point-in-time joins.
- Features can be shared and discovered across workspaces.
- Online stores support real-time applications.
- A free trial is listed.
What to know first
- Requires a Unity Catalog-enabled workspace.
- APIs cannot be called from non-Databricks environments.
- Feature Views and materialization are in Public Preview.
- Usage-based charges vary by product, cloud provider, and region.
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| Pay as you go | Monthly based on usage; per-second granularity. · No up-front costs · Pay for products used · Rates vary by product, cloud provider, and region | On request |
As docs.databricks.com lists them, read 30 Sept 2026. The maker's own page has the final word.
Getting started
- Open the Databricks Feature Store documentation.
- Use a Databricks workspace enabled for Unity Catalog.
- Start the available free trial.
- Install the Feature Engineering Python client from PyPI, or use its pre-installed version in Databricks Runtime 13.3 LTS ML and above.
- Author features as Feature Views through Databricks pipelines or as feature tables in Delta tables.
Verdict
Databricks Feature Store connects feature management with model lineage and serving, but it depends on a Unity Catalog-enabled workspace. Review the preview status, API environment limitation, and usage-based pricing components against your setup.
Questions
Is Databricks Feature Store free?
It is a paid service with a free trial. Pay as you go is billed monthly based on usage; rates vary by product, cloud provider, and region.
Which platforms does it support?
The listed platforms are API and web. Its Feature Store APIs cannot be called from local or other non-Databricks environments, though local IDE development and unit testing are supported.
Does it support online and offline stores?
Yes. It also supports feature monitoring, point-in-time joins, and both batch and real-time serving.
Which third-party online stores are supported?
Supported stores include Amazon DynamoDB, Amazon Aurora (MySQL-compatible), and Amazon RDS MySQL. Availability differs by Feature Store mode.
What does it cost to serve features?
Materialization uses serverless compute, feature serving endpoints use the Model Serving SKU, and online stores use Lakebase compute priced by capacity units and replicas.
Who makes Databricks Feature Store?
Databricks is headquartered in San Francisco, California, United States.
Facts from Databricks Feature Store's own pages
- Client limitation
- Feature Engineering and Feature Store APIs cannot be called from local or other non-Databricks environments, although local IDE development and unit testing are supported. · docs.databricks.com, 30 Sept 2026
- Feature authoring
- Features can be authored as declarative Feature Views managed through Databricks pipelines or as feature tables populated by writing values to Delta tables. · docs.databricks.com, 30 Sept 2026
- Feature Views status
- Feature Views and their feature materialization capability are in Public Preview. · docs.databricks.com, 30 Sept 2026
- Governance and discovery
- Registering features and models in Unity Catalog provides governance, lineage, point-in-time joins, and cross-workspace feature sharing and discovery. · docs.databricks.com, 30 Sept 2026
- Headquarters
- San Francisco, California, United States · docs.databricks.com, 28 Sept 2026
- Integrations
- Supported third-party online stores include Amazon DynamoDB, Amazon Aurora (MySQL-compatible), and Amazon RDS MySQL, with availability differing by Feature Store mode. · docs.databricks.com, 30 Sept 2026
- Model training and inference
- Models trained with Feature Store features automatically track feature lineage and look up the latest feature values at inference time. · docs.databricks.com, 30 Sept 2026
- Pricing components
- Materialization uses serverless compute, feature serving endpoints use the Model Serving SKU, and online stores use Lakebase compute priced by capacity units and replicas. · docs.databricks.com, 30 Sept 2026
- Pricing model
- Feature Store is billed at cost for underlying serverless compute, online store, and serving infrastructure, with no premium on top. · docs.databricks.com, 30 Sept 2026
- Purpose
- Databricks Feature Store is a central registry for the features used in AI and ML models. · docs.databricks.com, 30 Sept 2026
- Python client
- The Feature Engineering Python client is distributed as databricks-feature-engineering on PyPI and is pre-installed in Databricks Runtime 13.3 LTS ML and above. · docs.databricks.com, 30 Sept 2026
- Real-time serving
- Feature serving endpoints and online stores support serving features for real-time applications, with the main overview describing endpoints as providing millisecond latency. · docs.databricks.com, 30 Sept 2026
- Security and access
- Access to Unity Catalog feature tables is managed through Unity Catalog access controls. · docs.databricks.com, 30 Sept 2026
- Training and serving consistency
- Databricks says using the same feature computations at inference as during training eliminates training/serving skew. · docs.databricks.com, 30 Sept 2026
- Workspace requirement
- The current Databricks Feature Store requires a workspace enabled for Unity Catalog. · docs.databricks.com, 30 Sept 2026
Where Databricks Feature Store stands on Freedom251
- Feature Store Software#7 of 20