Hopsworks

What is Hopsworks?

Hopsworks is an AI platform that combines a feature store, a lakehouse and MLOps tooling for real-time machine learning, with the tagline The AI Lakehouse for Real-Time ML. It is used by ML teams building production systems, especially in financial services, retail and e-commerce, and government and defense. Named customers include Zalando, Paddy Power, Ericsson and Saab.

Key capabilities of Hopsworks

• Feature Store: stores and serves features with sub-millisecond retrieval powered by RonDB.
• AI Lakehouse: works with Delta, Iceberg and Hudi tables through a Python-native query engine.
• MLOps: experiment tracking, a model registry and deployment pipelines.
• Compute management: GPU management with Ray for training and KServe and vLLM for serving.
• Sovereign deployment: air-gapped, on-premises or hybrid.

How Hopsworks works

Teams compute features once with Spark, Flink, Pandas or DuckDB, store them in the feature store and serve them to models in training and in production.

Who uses Hopsworks?

AI and ML teams that need low-latency feature serving and control over where data lives.

Hopsworks pricing

Hopsworks offers Start Free signups, and prices are listed on a separate pricing page. No tier details are given on the main page.

Hopsworks alternatives

ZenML is an open-source framework for ML pipelines. Dataiku is a broad data science and analytics platform. DataRobot focuses on automated ML and AI lifecycle management.