- Home
- Model Registry Software
- Hopsworks AI Lakehouse
Hopsworks AI Lakehouse
Stockholm, Sweden
What it is
Hopsworks AI Lakehouse is a platform for building batch, real-time, and LLM AI systems on open standards. It supports Apache Iceberg, Delta Lake, and Apache Hudi, and its feature pipelines connect to a low-latency serving database with sub-second feature freshness. The Query Service transfers data to Python clients using Arrow and supports temporal joins, pushdown filters, credential vending, and reproducible training data creation. LLM workflows can query Lakehouse data from prompts, use vector search for retrieval-augmented generation, and create instruction datasets for fine-tuning while maintaining data governance. Feature, training, and inference pipelines cover batch, streaming, real-time, and LLM work, with trained models sent to a model registry. Integrations include Airflow, Apache Flink, Databricks, Snowflake, AWS SageMaker, Spark, Vertex AI, and Weights & Biases. Deployment options include managed SaaS, Kubernetes on AWS, Azure, or GCP, and air-gapped data centres. The platform lists encryption, project access controls, audit logs, 2FA, and SSO. It offers a free tier and is available through API, Linux, self-hosted, and web options.
For and against
For
- Supports three open table formats.
- Includes feature, training, and inference pipelines.
- Offers managed SaaS, Kubernetes, and air-gapped deployment.
- Free plan includes one project and community support.
Against
- Kubernetes installation requires version 1.27 or higher.
- Kubernetes deployment recommends at least 4–5 nodes.
- Enterprise pricing is custom.
The verdict
Hopsworks AI Lakehouse brings data, feature, and model workflows together across several deployment types. Teams should consider its Kubernetes requirements and the distinction between community and dedicated enterprise support.
Who it suits
The platform is aimed at developers, data scientists, platform and security engineers, and administrators building AI systems. It may suit teams needing batch, streaming, or LLM workflows and flexible deployment options.
Plans and prices
Free
1 project · Feature Store + Model Registry · Community support · No credit card required
Read 3 Oct 2026 · hopsworks.ai
Enterprise
All SaaS features · On-premise & air-gapped · Dedicated support team · Custom integrations
Read 3 Oct 2026 · hopsworks.ai
SaaS
Unlimited projects · Feature Store + Model Registry · Model Serving · Usage-based pricing
Read 3 Oct 2026 · hopsworks.ai
On the record
- AI pipelines
- Its feature, training, and inference pipelines support batch, streaming, real-time, and LLM AI workflows, with trained models output to a model registry.hopsworks.ai · read 3 Oct 2026
- Compliance
- The security page lists ISO 27001 and SOC 2 Type 2 certification and states that Hopsworks is GDPR compliant.hopsworks.ai · read 3 Oct 2026
- Deployment
- The documentation describes managed SaaS and deployments on managed Kubernetes in AWS, Azure, or GCP, as well as air-gapped data centres.docs.hopsworks.ai · read 3 Oct 2026
- Deployment requirements
- The Kubernetes installer page lists Kubernetes 1.27 or higher and recommends a minimum of 4–5 nodes.hopsworks.ai · read 3 Oct 2026
- Integrations
- Listed integrations include Airflow, Apache Flink, Databricks, Snowflake, AWS SageMaker, Spark, Vertex AI, and Weights & Biases.hopsworks.ai · read 3 Oct 2026
- Intended users
- The documentation identifies developers, data scientists, platform engineers, security engineers, and administrators as user paths for the platform.docs.hopsworks.ai · read 3 Oct 2026
- LLM workflows
- The platform supports querying Lakehouse data from LLM prompts, vector search integration for RAG, and instruction dataset creation for fine-tuning while maintaining data governance.hopsworks.ai · read 3 Oct 2026
- Purpose
- Hopsworks describes its AI Lakehouse as a platform for building batch, real-time, and LLM AI systems on open standards.hopsworks.ai · read 3 Oct 2026
- Python query service
- Hopsworks Query Service uses Arrow-native transfer to Python clients and supports temporal joins, pushdown filters, credential vending, and reproducible training data creation.hopsworks.ai · read 3 Oct 2026
- Real-time features
- The product page says streaming feature pipelines connect to a low-latency serving database and provide sub-second feature freshness.hopsworks.ai · read 3 Oct 2026
- Security
- Hopsworks says it encrypts data at rest and in transit and provides project-based access controls, audit logs, and authentication options including 2FA and SSO.hopsworks.ai · read 3 Oct 2026
- Support
- The Free and SaaS pricing tiers include community support, while Enterprise includes a dedicated support team and guaranteed SLA.hopsworks.ai · read 3 Oct 2026
- Table formats
- It natively supports Apache Iceberg, Delta Lake, and Apache Hudi.hopsworks.ai · read 3 Oct 2026
