Skip to main content

Getting Started

What You'll Build​

In this quickstart, you'll use the OpenAI GPT-4o model through the Orchestration Service of AI Core for generating text. The application will send a prompt to the AI model and display the generated response.

Prerequisites​

This quickstart assumes you are using the default resource group of your AI Core instance. If you want or need to use a different resource group, please ensure that resource group has an orchestration deployment available.

Add the SDK as a Dependency​

Add the following dependency to your pom.xml file:

<dependency>
    <groupId>com.sap.ai.sdk</groupId>
    <artifactId>orchestration</artifactId>
    <!-- Use the latest version here -->
    <version>${ai-sdk.version}</version>
</dependency>

Use the Orchestration API​

We'll use a client to interact with the Orchestration service:

var client = new OrchestrationClient();

Next, we'll specify the model we want to use:

var config = new OrchestrationModuleConfig()
        .withLlmConfig(OrchestrationAiModel.GPT_4O);

Now we can create our first prompt:

var prompt = new OrchestrationPrompt("Hello world! Why is this phrase so famous?");

var result = client.chatCompletion(prompt, config).getContent();

The result will be the text generated by the AI model.

Run the Application Locally​

In order to run the application locally, you need to provide credentials for your AI Core service instance.

For this example we'll use a service key and pass it as an environment variable to the application.

cd your-spring-app/

# assuming a bash, for other shells (e.g. PowerShell) see the below documentation
export AICORE_SERVICE_KEY='{ "clientid": "...", "clientsecret": "...", "url": "...", "serviceurls": { "AI_API_URL": "..." } }'

# assuming Maven and a Spring Boot application
mvn spring-boot:run

Please find detailed instructions and more examples in this documentation.

Explore Further Capabilities​

Check out the options available for the OrchestrationPrompt and OrchestrationModuleConfig classes. You can use templating, content filtering, data masking and more. Please refer to this documentation for more information.