api-spring-ai-prompt-openai-v1

Spring Boot and Spring AI example that uses prompt engineering with OpenAI. It shows how to combine a system role, user instructions and a structured output format to turn a short business request into a requirements analysis. It also exposes a REST endpoint that forwards a question to the model.

This project is part of the Java & Spring AI Examples series (module 02 – Prompt Engineering).

Project page: https://josesaid.github.io/api-spring-ai-prompt-openai-v1/

What it does

On startup, PromptRunner runs two CommandLineRunner beans that call the model through ChatClient:

Requirements analysis: a system role (business analyst) with explicit rules, a business request and a required Markdown structure (objective, actors, functional and non-functional requirements, business rules and missing information).
Dynamic role prompt: a second role (application architect) analyzes a security incident-management request.

The REST endpoint sends any question to the same model through ChatService.

Tech stack

Component	Detail
Java	24 (see `pom.xml`)
Spring Boot	4.x (Spring Boot parent in `pom.xml`)
Spring AI	2.0.0
Model provider	OpenAI (chat model, profile `openai`)
Build tool	Maven (wrapper included)

Project structure

src/main/java/com/mx/development/jose/said/
├── ApiSpringAiDemoV1Application.java   # Spring Boot entry point
├── controller/ChatController.java      # GET /api/v1/chat-service/chat
├── service/ChatService.java            # Service contract
├── service/ChatServiceImpl.java        # Calls the model through ChatClient
└── runner/PromptRunner.java            # Startup prompt examples
src/main/resources/
├── application.yml                     # Port, active profile, logging
└── application-openai.yml              # OpenAI configuration

Prerequisites

JDK 24, as declared in pom.xml
Maven (the Maven Wrapper is included, so a local installation is optional)
An OpenAI API key, created at https://platform.openai.com/api-keys

Configuration

The API key is read from the API_KEY environment variable. Never commit a real key to the repository.

export API_KEY=your_openai_api_key

In PowerShell:

$Env:API_KEY = "your_openai_api_key"

Running the application

./mvnw spring-boot:run

The application starts on port 8088. The two prompt examples print their responses to the console at startup.

REST API

Method	Path	Query parameter	Description
GET	`/api/v1/chat-service/chat`	`question`	Sends the question to the model

curl "http://localhost:8088/api/v1/chat-service/chat?question=Explain%20prompt%20engineering%20in%20one%20paragraph"

Key concepts

System role: sets the model's behavior and rules.
User message: carries the task and its input data.
Structured output: a fixed set of Markdown sections makes the answer predictable.
Explicit constraints: rules such as "do not invent information" reduce hallucinations.

Author

Created by Jose Said Olano Garcia.