You are viewing the RapidMiner Hub documentation for version 10.0 - Check here for latest version
RapidMiner Real-Time Scoring
RapidMiner Real-Time Scoring is an add-on product to RapidMiner Server designed for fast scoring use cases via web services. Its components are called Scoring Agents. They have a minimal memory footprint and fast response times. Each Scoring Agent requires a separate license.
Scoring Agents can be used as web services and serve pre-defined deployments which can run anywhere, even on Edge devices as long as those support the minimum system requirements.
Once you have designed your process in RapidMiner Studio, trained it, and validated it, it's the moment to put it to work and use it to predict something.
- If that prediction is part of a scheduled batch process, then RapidMiner Server will help you.
Although the word "scoring" is often used in risk management, any process can be deployed in the Scoring Agent, and should be deployed in the Scoring Agent if the emphasis is on a web application with high throughput -- fast predictions and low latency.
Find more information about:
- System requirements
- How to install a Scoring Agent
- How to create a deployment from RapidMiner AI Hub or Projects
- How to install a deployment
- How to score data
- How to update a deployment
- How to upgrade a Scoring Agent
REST API
The complete REST API documentation of the Scoring Agent is available as OpenApi 3.0 specification and is
published on SwaggerHub.