The SAP Cloud SDK for AI is the official Software Development Kit (SDK) for SAP AI Core, SAP Generative AI Hub, and Orchestration Service. It provides Java developers with a comprehensive toolkit for integrating AI capabilities into business applications running on SAP Business Technology Platform (BTP). The SDK abstracts the complexity of connecting to AI services, managing deployments, and executing AI operations through a type-safe, idiomatic Java API. pom.xml9-11 README.md9-13
This SDK supports multiple integration patterns: direct foundation model access via the OpenAI client, sophisticated AI pipelines through the Orchestration Service, and core AI Core platform operations for deployment and lifecycle management. For detailed information on specific modules, see Project Structure, Key Features, and Model Lifecycle & Support.
Sources: README.md7-13 pom.xml9
The SDK is organized as a multi-module Maven project with a parent POM at pom.xml1-101 defining dependency versions, plugin configurations, and quality standards. The project consists of several published modules plus sample code:
SDK Module Dependency Graph Title: SDK Module Dependency Graph
Maven Module Configuration
| Module | Artifact ID | Group ID | Description |
|---|---|---|---|
| Parent | sdk-parent | com.sap.ai.sdk | Root POM with dependency management pom.xml5-9 |
| Core | core | com.sap.ai.sdk | AI Core client and connectivity infrastructure core/pom.xml9-11 |
| Orchestration | orchestration | com.sap.ai.sdk | Orchestration Service with LLM pipeline features orchestration/pom.xml9-11 |
| OpenAI | openai | com.sap.ai.sdk.foundationmodels | Azure OpenAI foundation models client foundation-models/openai/pom.xml10-13 |
| SAP RPT | sap-rpt | com.sap.ai.sdk.foundationmodels | SAP RPT table completion models foundation-models/sap-rpt/pom.xml11-13 |
| Grounding | document-grounding | com.sap.ai.sdk | Document grounding service client core-services/document-grounding/pom.xml10-12 |
| Prompt Registry | prompt-registry | com.sap.ai.sdk | Prompt lifecycle management client core-services/prompt-registry/pom.xml10-14 |
| Batch | batch | com.sap.ai.sdk | Asynchronous batch processing service pom.xml34 |
| Spring App | spring-app | com.sap.ai.sdk.app | Sample application and E2E tests sample-code/spring-app/pom.xml10-13 |
For details, see Project Structure.
Sources: pom.xml31-39 core/pom.xml9-11 orchestration/pom.xml9-11 foundation-models/openai/pom.xml10-13 core-services/document-grounding/pom.xml10-12 core-services/prompt-registry/pom.xml10-14 sample-code/spring-app/pom.xml10-13
The SDK provides high-level abstractions for common AI tasks, primarily through the Orchestration and Foundation Model modules.
OrchestrationClient README.md89-106DpiMasking configurations README.md131document-grounding module core-services/document-grounding/pom.xml12batch module sample-code/spring-app/pom.xml103For details, see Key Features.
Sources: README.md11-13 README.md89-106 core-services/document-grounding/pom.xml12 sample-code/spring-app/pom.xml103
The SDK maintains a typed list of supported AI models across different providers. It tracks the lifecycle of these models, including deprecation and retirement dates.
Model Entity Mapping Title: Model Entity Mapping
For details, see Model Lifecycle & Support.
Sources: foundation-models/openai/pom.xml10-13 orchestration/pom.xml9-11
The SDK bridges the gap between Java application code and the SAP AI Core infrastructure using the SAP Cloud SDK's connectivity features.
Connectivity Logic Flow Title: Connectivity Logic Flow
Authentication and Configuration
Developers can configure connectivity using environment variables like AICORE_SERVICE_KEY README.md120 The SDK handles the underlying OAuth2 flows and destination resolution automatically via SAP Cloud SDK modules core/pom.xml49-64
Sources: README.md112-127 core/pom.xml49-64 orchestration/pom.xml55-61
Refresh this wiki