Hopsworks
Best for self-hosting
Why teams pick it
Hopsworks delivers a real-time AI lakehouse, offering a Python-focused feature store, MLOps tools, multi-tenant projects, and flexible deployment on cloud, serverless, or on-premises.
Open-source alternatives to Databricks Feature Store
Compare community-driven replacements for Databricks Feature Store in feature stores workflows. We curate active, self-hostable options with transparent licensing so you can evaluate the right fit quickly.
Databricks Feature Store
Databricks Feature Store secures features with ACLs, tracks full lineage, and integrates with MLflow and Unity Catalog for consistent training and online inference.Read more
Key stats
- 5Alternatives
- 1Support self-hosting
Run on infrastructure you control
- 2Active development
Recent commits in the last 6 months
- 4Permissive licenses
MIT, Apache, and similar licenses
Counts reflect projects currently indexed as alternatives to Databricks Feature Store.
Start with these picks
These projects match the most common migration paths for teams replacing Databricks Feature Store.
Featureform
Privacy-first alternative
Why teams pick it
Keep customer data in-house with privacy-focused tooling.
All open-source alternatives
Hopsworks
Real-time AI Lakehouse with Python-centric Feature Store
Self-host friendlyIntegration-friendlyAI-powered workflowsJava
Why teams choose it
- Python‑centric Feature Store with versioning, lineage, and governance
- Project‑based multi‑tenant environment for team collaboration
- Integrated MLOps stack: Airflow, Jupyter, GPU training, model serving
Watch for
On‑premises install requires minimum 32 GB RAM, 8 CPUs
Migration highlight
Fraud detection with real-time scoring
Features are ingested and served instantly, enabling online models to flag fraudulent transactions as they occur.
Feathr
Scalable feature store for unified data and AI engineering
Permissive licenseFast to deployIntegration-friendlyScala
Why teams choose it
- Pythonic APIs with native PySpark and Spark SQL UDF support
- Point-in-time correct feature computation for training and online serving
- Scalable architecture handling billions of rows and petabyte-scale data
Watch for
Requires a Spark environment, limiting use with non‑Spark stacks
Migration highlight
NYC Taxi fare prediction
Rapidly define, materialize, and serve fare prediction features with point‑in‑time correctness
OpenMLDB
SQL‑driven feature platform delivering millisecond real‑time ML features
Active developmentPermissive licenseIntegration-friendlyC++
Why teams choose it
- Consistent feature generation for training and inference via unified execution plan
- Real‑time SQL engine produces features in a few milliseconds
- SQL‑first feature definition with extensions like LAST JOIN and WINDOW UNION
Watch for
Requires SQL expertise; non‑SQL environments may need adaptation
Migration highlight
NYC Taxi Trip Duration Prediction
End‑to‑end ML pipeline built with OpenMLDB and LightGBM to predict ride duration, demonstrating rapid feature development and deployment.
Featureform
Turn existing data pipelines into a collaborative virtual feature store
Permissive licensePrivacy-firstFast to deployGo
Why teams choose it
- Infrastructure-agnostic: works with existing data platforms
- Immutable feature definitions with lineage and versioning
- Built-in RBAC, audit logs, and dynamic serving rules
Watch for
Requires orchestration setup and configuration
Migration highlight
Notebook-to-Production Feature Pipeline
Data scientists push transformations from Jupyter notebooks to a central repository, then Featureform orchestrates Spark jobs and serves the feature in Redis for online inference.
Feast
Unified feature store for training and real‑time inference
Active developmentPermissive licenseIntegration-friendlyPython
Why teams choose it
- Consistent feature availability for training and serving
- Point‑in‑time correct datasets to prevent leakage
- Single data access layer that abstracts storage backends
Watch for
Web UI is experimental and may lack full functionality
Migration highlight
Batch model training with historical features
Generate point‑in‑time training datasets that avoid data leakage.
Choosing a feature stores alternative
Teams replacing Databricks Feature Store in feature stores workflows typically weigh self-hosting needs, integration coverage, and licensing obligations.
- 1 project let you self-host and keep customer data on infrastructure you control.
- 2 options are actively maintained with recent commits.
Tip: shortlist one hosted and one self-hosted option so stakeholders can compare trade-offs before migrating away from Databricks Feature Store.