Let’s talk fast, accurate AI at AWS re:Invent.

Join us in Vegas on Dec. 1-5.
Feature Store

Feature store

Feature store
A virtual feature store for the modern AI stack

Redis’ acquisition of Featureform makes Redis a complete end-to-end feature store. It combines our speed for online serving with orchestration and governance to deliver a unified stack for real-time, batch, and streaming workflows. Teams can use Redis as a full feature store solution or as the online store with tools like Feast and Tecton.

Feature store
Serve real-time features to models in production

Serve real-time features to models in production

Deliver live features, like user behavior or risk scores, to your models with sub-millisecond latency. Our feature store orchestrates batch, streaming, and on-demand pipelines while serving them instantly at inference.

Faster feature iteration with reusable pipelines

Faster feature iteration with reusable pipelines

Define features once and reuse them across training and inference. Versioning, lineage, and search make iteration fast, consistent, and collaborative.

Integrate with your existing stack—fast

Integrate with your existing stack—fast

We can serve as your complete feature store solution with no rewrites or platform lock-in. You can also integrate with frameworks like Feast and Tecton if you only need Redis for online serving.

Built for enterprise scale and control

Built for enterprise scale and control

Deploy on your cloud or on-prem with robust support for access control, auditability, and team-level collaboration. Our governance features help enterprises stay compliant without compromising speed.

ifood
<1msmillisecond per read
DoorDash
38%decrease in latency
"

It’s not just about putting in a food order. It’s about providing an optimal experience for our customers. It’s really, really fast. Plus it’s a lot cheaper.

Daniel GalinkinHead of ML Engineering
"

We also saw a 38% decrease in Redis latencies, helping to improve the runtime performance of serving models.

Arbaz KhanStaff Software Engineer
  • ifood
  • DoorDash

Redis Feature Store (formerly Featureform)

Search (32px)

Versioning, lineage, & discovery

Track feature versions, metadata, and ownership to support fast iteration and easy collaboration.

Get the details
Tech Diagrams

Iceberg-native architecture

Scale with Apache Iceberg for indexing, time travel, and high-performance access to large datasets.

Learn about Iceberg
Dashboard

Point-in-time correctness

Train with historical feature views and avoid data leakage.

Take a closer look
Monitoring

Governance & monitoring

Manage access, enforce policies, and monitor feature health across teams and environments.

Check out our docs

Redis as an online store

Integrated modules

Integrate with Feast & other feature stores

Use Feast as your offline feature store and Redis for real-time online serving.

See how it works
Forecasting

Ultra-low-latency predictions

Get sub-millisecond performance for use cases like fraud detection, recommendations, and dynamic pricing.

Watch a demo
High Availability

Reduce downtime

Five-nines availability and single-digit-second failover maintain reliability and protect data integrity for mission-critical AI apps.

Learn more
Real time analytics

Keep features fresh in real time

Feature values update instantly, so models always access the most recent data for accurate predictions and dynamic decision-making.

Read our blog

Redis Open Source vs. Enterpise-grade Redis

Here is a breakdown of the additional capabilities in our enterprise-grade offering

Capability
Open source
Enterprise grade
Feature definitions & registry
Feature versioning & lineage
Batch ingestion
Online feature serving
Vector & embedding support
Deployment flexibility (cloud/on-prem)
Streaming ingestion
Security & access control (RBAC, SSO/SAML)
Monitoring & governance
Enterprise support & SLAs

Get started

Meet with a Redis expert and build your feature store today.