logicalclocks/hopsworks

Health v2 · maintenance only

Adoption not part of the score

1303 stars · 160 forks observed · 2026-08-28

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

Hopsworks is an open-source, data-intensive AI platform (an 'AI Lakehouse') built around a Python-centric Feature Store with online/offline storage powered by RonDB, plus MLOps capabilities including a model registry, model serving, and experiment tracking. It is a modular Java-based platform deployable on Kubernetes (AWS, Azure, GCP, on-prem) that integrates with Spark, Flink, Pandas, Databricks, SageMaker, and Kubeflow.

Use cases

• manage and reuse ML features across models with a feature store
• serve features with sub-millisecond online latency for real-time inference
• prevent training-serving skew between batch and online data
• track experiments and manage a model registry
• deploy and monitor ML models in production
• build feature pipelines with Spark, Flink, or Pandas
• govern and share ML assets across data science teams
• run an on-premise or air-gapped ML platform

When to choose

• you need an open-source feature store with both online and offline stores
• your team requires feature reuse, governance, and lineage for ML data
• you want an integrated MLOps platform with model registry and serving
• you need sub-millisecond feature vector retrieval at serving time
• you deploy on Kubernetes across AWS, Azure, GCP, or on-premises

When to avoid

• you only need a lightweight feature store without a full platform (consider Feast)
• you cannot accept the AGPL-3.0 license for your use case
• you lack Kubernetes or multi-node infrastructure for deployment
• your ML stack is simple enough that a full lakehouse platform is overkill

Facets

application · maturity active

machine-learning data-science etl streaming monitoring search-engine vector-database self-hosted workflow-automation sdk machine-learning data-science large-language-models big-data databases cloud self-hosted python jvm feature-store mlops model-registry model-serving ai-lakehouse rondb online-offline-store training-serving-skew kserve spark flink pandas data-governance agpl data-engineering kubernetes docker web-server

Member repositories

For agents

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