RedisAI

AI serving engine for real-time applications

 
AI creates new data-layer challenges for organizations, including handling the proliferation and complexity of inference data. Requiring no added infrastructure, RedisAI allows you to run your inference engine where the data lives, decreasing latency and increasing simplicity—all coupled with the core Redis Enterprise features.

 

Introduction to RedisAI from RedisConf 2020 keynote

Benefits

AI inferencing where your data lives

Enriching AI transactions with reference data can be vastly slower than the AI processing itself. RedisAI minimizes latency by retrieving the reference data directly from the database’s shared memory.

Deploy new models with no downtime or performance penalties

Update models transparently without affecting inference performance. Fully integrated with MLFlow for managing your AI lifecycle.

Serve AI over a robust, scalable, and production-proven platform

With RedisAI you also get all of Redis Enterprise’s features, including high availability (99.999%) and infinite linear scalability without performance trade-offs.

Superior performance

Delivering up to 9x more throughput than other AI model-serving platforms.

Built-in support for all major AI backends

Serve machine-learning and deep-learning models trained by state-of-the-art platforms like TensorFlow, PyTorch, or ONNXRuntime. Run inferences across platforms.

Runs everywhere

Run on CPUs, state-of-the-art GPUs, high-end compute engines, or even tiny Raspberry Pi or NVIDIA Jetson devices.

Use cases

Transaction scoring and fraud detection

Reduce fraud with real-time transaction scoring. Enhanced credit-decision making with reference data at blazing speed.

Recommendation engine and personalization

AI-powered retail analytics and decisions. Drive more revenue with better personalization and AI-powered product recommendations.

Anomaly detection

RedisAI can leverage time-series data in RedisTimeSeries for forecasting and anomaly detection.

Increased search relevancy

Combine RedisAI with RediSearch to increase search relevancy and create better user experiences.

Power knowledge graphs

Use your data to create knowledge graphs and serve them in RedisGraph. These knowledge graphs can give AI models context for better inferencing.


Supported backends

TensorFlow
PyTorch

ONNXRuntime