- Home
- Feature Store Software
- Red Hat OpenShift AI Feature Store
Red Hat OpenShift AI Feature Store
Raleigh, North Carolina, United States
What it is
Red Hat OpenShift AI Feature Store provides reusable machine-learning features for model training and inference. Teams can push features into the store and retrieve them when needed. It is based on the open-source Feast project and runs within self-managed OpenShift AI, which requires Red Hat OpenShift. A cluster administrator enables the capability through the OpenShift AI Operator and creates an instance in a data science project. The default local provider uses a SQL registry or local file, a Parquet offline store, and a SQLite online store. The online container exposes a Feature Store REST API inside the OpenShift cluster. Documentation also describes PostgreSQL for the online store and S3 for the registry. The feature store supports online and offline stores, point-in-time joins, feature monitoring, and both serving modes. Red Hat positions OpenShift AI for data scientists, AI engineers, developers, and IT operations teams, with deployment across on-premise, edge, and disconnected environments. Pricing is on request, and a 60-day self-supported trial is available.
For and against
For
- Reusable features for training and inference
- Online and offline stores
- REST API inside the cluster
- Point-in-time joins and monitoring
- 60-day self-supported trial
Against
- Requires Red Hat OpenShift
- Requires administrator setup
- Pricing is on request
- Trial is self-supported
The verdict
OpenShift AI Feature Store gives teams Feast-based feature management within a self-managed Red Hat environment. OpenShift dependency, administrator setup, and request-based pricing are central considerations.
Who it suits
This feature store suits data science and IT teams operating self-managed OpenShift AI environments. It is relevant when reusable features, online/offline serving, and private deployment are needed.
Plans and prices
Red Hat OpenShift AI Self-Managed
Core-pair or bare-metal subscriptions · Standard (8x5) or Premium (24x7) support · Requires Red Hat OpenShift
Read 7 Oct 2026 · redhat.com
On the record
- API
- The online container makes a Feature Store REST API available inside the OpenShift cluster.docs.redhat.com · read 7 Oct 2026
- Default stores
- The local provider uses a SQL registry or local file, a Parquet offline store, and SQLite online store.docs.redhat.com · read 7 Oct 2026
- Intended users
- Red Hat describes OpenShift AI as a platform for data scientists, AI engineers, developers, and IT operations teams.redhat.com · read 7 Oct 2026
- Platform dependency
- OpenShift AI is layered on Red Hat OpenShift, which must be installed before self-managed OpenShift AI.redhat.com · read 7 Oct 2026
- Purpose
- Feature Store provides reusable machine-learning features that teams can push and retrieve for model training and inference.docs.redhat.com · read 7 Oct 2026
- Security
- The product page describes deploying models across on-premise, edge, and disconnected environments to support private AI.redhat.com · read 7 Oct 2026
- Security certification
- Red Hat announced that OpenShift AI achieved an initial ISO/IEC 42001 AI management system certification audit.access.redhat.com · read 7 Oct 2026
- Setup
- A cluster administrator enables Feature Store through the OpenShift AI Operator and creates an instance in a data science project.docs.redhat.com · read 7 Oct 2026
- Storage integrations
- The documentation describes configuring a PostgreSQL online store and an S3 registry, and points to Feast for other store options.docs.redhat.com · read 7 Oct 2026
- Support
- OpenShift AI subscriptions are offered with Standard (8x5) or Premium (24x7) support SLAs.redhat.com · read 7 Oct 2026
- Trial
- Red Hat offers a 60-day self-supported OpenShift AI Self-Managed trial; the trial page also lists an OpenShift Container Platform trial for users without a cluster.redhat.com · read 7 Oct 2026
- Underlying project
- OpenShift AI Feature Store is based on the open-source Feast project.docs.redhat.com · read 7 Oct 2026
