OpenAI: Chat integration with Spring Boot

This repository demonstrates how to integrate OpenAI chat functionality into a Spring Boot project.

1. Create a Spring Boot Project

First, create a new Spring Boot project using your preferred method (e.g., Spring Initializr or an IDE like IntelliJ IDEA or Eclipse). Make sure your project is set up with Maven.

2. Add dependencies to the pom.xml file:

<!--	Spring Web	-->
<dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-web</artifactId>
</dependency>

<dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-test</artifactId>
    <scope>test</scope>
</dependency>

<!--	Lombok	-->
<dependency>
    <groupId>org.projectlombok</groupId>
    <artifactId>lombok</artifactId>
    <optional>true</optional>
</dependency>

<!--	WebFlux	-->
<dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-webflux</artifactId>
</dependency>

<!--	JSON parsing	-->
<dependency>
    <groupId>com.fasterxml.jackson.core</groupId>
    <artifactId>jackson-databind</artifactId>
</dependency>

3. Define your secrets in application.yml

openai:
  api-key: ${OPENAI_TOKEN}
  assistant:
    name: ${ASSISTANT_NAME}
    model: ${ASSISTANT_MODEL}
    instructions: ${ASSISTANT_INSTRUCTIONS}

4. Create an OpenAI configuration class:

@Configuration
public class OpenAiConfig {

    private final String apiKey;

    public OpenAiConfig(@Value("${openai.api-key}") String apiKey) {
        this.apiKey = apiKey;
    }

    @Bean
    public WebClient openAiWebClient() {
        return WebClient.builder()
                .baseUrl("https://api.openai.com/v1")
                .defaultHeader(HttpHeaders.AUTHORIZATION, "Bearer " + apiKey)
                .defaultHeader(HttpHeaders.CONTENT_TYPE, MediaType.APPLICATION_JSON_VALUE)
                .defaultHeader("OpenAI-Beta", "assistants=v2")
                .build();
    }
}

---

Now that we’ve completed the configuration, we can move on to explaining the available endpoints.

⚙️ Assistant Controller

POST /assistant/new

Sends a request to create a new assistant using the configuration defined in application.yml.

curl --location --request POST 'http://localhost:8080/assistant/new'

Response:

ID: asst_123abc456def...

GET /assistant/{assistantId}

Returns the details of an existing assistant by ID.

curl --location 'http://localhost:8080/assistant/asst_js0TkCNlOgIxqt6ir3gGmt2O'

POST /assistant/{assistantId}/upload

Uploads a file and attaches it to the specified assistant.
The file is also stored in a vector store and linked to the assistant for retrieval and search purposes.

curl --location 'http://localhost:8080/assistant/asst_js0TkCNlOgIxqt6ir3gGmt2O/upload' \
--form 'file=@"C:/Users/Documents/data.json"'

Response:

{
  "fileId": "file_abc123...",
  "vectorStoreId": "vs_xyz456..."
}

GET /assistant/{assistantId}/files

Lists all files attached to the assistant.

curl --location 'http://localhost:8080/assistant/asst_js0TkCNlOgIxqt6ir3gGmt2O/files' \
--header 'OpenAI-Beta: assistants=v2'

⚙️ Thread Controller

POST /threads/chat

Starts or continues a conversation thread between the user and an assistant.
If a threadId is not provided, a new thread is created.
The message is added to the thread, the assistant processes it, and the response is returned after execution is complete.

curl --location 'http://localhost:8080/threads/chat' \
--data '{
    "assistantId": "asst_js0TkCNlOgIxqt6ir3gGmt2O",
    "threadId": "thread_rzSQwwxikoALUUlSQ35OIfcM",
    "message": "What are the contents of the uploaded file?"
}'

Request:

{
  "assistantId": "asst_js0TkCNlOgIxqt6ir3gGmt2O",
  "threadId": "",  // Optional: leave empty to create a new thread
  "message": "What are the contents of the uploaded file?"
}

Response:

{
  "threadId": "thread_xyz789...",
  "runId": "run_abc456...",
  "answer": "The file contains a list of user transactions from January 2024..."
}

Behavior:

• If threadId is omitted or blank, a new thread is automatically created.
• The message is sent to the assistant, which processes it.
• The system waits for the assistant’s response before replying to the client.

---

📁 Project Resources

The resources/assets directory contains the following files:

• data.json:
  An example input file containing structured data (in Brazilian Portuguese).

• PROJECT_OPENAI.postman_collection.json:
  A Postman collection with predefined requests to test the API endpoints.

• instructions.txt:
  A file with prompt instructions for the Assistant (written in Brazilian Portuguese).

---

🔗 Useful Links

• Spring Initializr - Quickly bootstrap a Spring Boot project.
• OpenAI Platform - Learn more about OpenAI and its developer platform.

---

Feel free to contribute by submitting pull requests or reporting issues.
Contributions are welcome! 🚀

Cheers,
Victor