Skip to main content

Querying Features and Models in production

Get feature values in production is done using the Core API over gRPC or the REST API.

Accessing the Core Kuberentes service​

Installing Raptor Core will create a Kubernetes Service that exposes the Core API.

Service name: raptor-core-service Namespace: raptor-system gRPC port: 60000 REST API port: 60001

The Core API​

The Core API allows you to access the features values and models predictions in a low-latency and high-throughput manner, as well as accessing the low-level operations of the engine(Such as setting/incrementing/

For more information about the Core API, please refer to the Reference section.

Accessing from production code​

The Core API is exposed over gRPC and REST, and can be accessed from any language that supports gRPC or REST.

Accessing Model predictions​

Models can be accessed using the same API for feature values. Just use the Model FQN as the selector 🙃