Building machine learning use cases becomes quick, easy, and more maintainable with AI Accelerator. Customize the default configuration of your use cases by updating the machine learning models and feature extractors that suit your requirements. Save the extracted features and scoring responses. Use the AI Accelerator Connect API to generate predictions in real time.
AI Accelerator Standard Objects
Use the standard objects to configure your use case by defining the feature extractor, the machine learning model, and the objects and fields for saving the prediction results for your use case.
Use the AIScoringStatusEvent event to notify subscribers after a scoring job is processed.
AI Accelerator Business REST APIs
Use the AI Accelerator Business APIs to get real-time score predictions for a machine learning use case model. You can access AI Accelerator APIs using REST endpoints. These REST APIs follow similar conventions as Connect REST APIs.
AI Accelerator Predictions API Using Apex
This AI Accelerator Predictions API is also exposed to Apex. You can access/invoke this API from within your Apex class to get realtime prediction results.
This Apex reference goes into detail about the built-in Apex ConnectApi classes and interfaces.
Use the AI Accelerator metadata types to set up your machine learning use case. Define machine learning models, feature extractors, and the objects and fields for saving the prediction results for your use case.
AI Accelerator Tooling Objects
Tooling API exposes metadata used in developer tooling that you can access through REST or SOAP. Tooling API’s SOQL capabilities for many metadata types allow you to retrieve smaller pieces of metadata. For more information about Tooling API objects and to find a complete reference of all the supported objects, see Introducing Tooling API.
j
