AdamAI – AI Integrated Spring Boot Service

AdamAI is a Spring Boot based middleware service that acts as a bridge between a user and a Large Language Model (LLM) using OpenRouter AI.

The application supports:

• AI personalities (helper, coder, pirate)
• Session memory/context handling
• Retry & exponential backoff for failed AI requests
• Swagger/OpenAPI documentation
• Global exception handling
• Frontend chat interface
• REST API communication

---

Features

AI Personalities

Users can choose between different personalities:

• helper
• coder
• pirate

Each personality maps to a different system prompt before sending requests to the AI model.

---

Session Memory

The application stores conversation history in-memory using:

Map<String, List<AIMessage>>

This allows the AI to remember previous messages in the same conversation.

---

Retry & Exponential Backoff

Failed AI requests automatically retry using Spring Retry.

Implemented with:

@Retryable(
        retryFor = Exception.class,
        maxAttempts = 3,
        backoff = @Backoff(
                delay = 1000,
                multiplier = 2
        )
)

Retry functionality is enabled globally using:

@EnableRetry

This improves resilience when the AI service is temporarily unavailable.

---

Global Exception Handling

The application uses:

@ControllerAdvice

to return clean and user-friendly error responses.

---

Swagger / OpenAPI

API documentation is automatically generated using SpringDoc OpenAPI.

Swagger UI:

http://localhost:8080/swagger-ui.html

---

Technologies Used

• Java 25
• Spring Boot 4
• Spring Web
• Spring Validation
• Spring Retry
• OpenRouter AI
• Swagger / OpenAPI
• JUnit 5
• Mockito
• MockMvc
• JaCoCo
• HTML / CSS / JavaScript

---

Project Structure

src
├── client
├── controllers
├── DTO
├── exceptions
├── services
├── resources
│   ├── static
│   └── application.properties
└── test

---

API Endpoint

POST /api/v1/chat

Request Body

{
  "message": "Hello AI",
  "personality": "helper",
  "sessionId": "user-123"
}

Response

{
  "reply": "Hello! How can I help you today?"
}

---

Validation

The API validates incoming request bodies before processing them.

The application validates incoming requests using:

@NotBlank

If validation fails, the API returns:

400 Bad Request

---

Testing

The project includes:

• Controller tests using MockMvc
• Service layer unit tests
• Validation tests
• Global exception handler tests
• AIClient personality tests
• Mocked dependencies using Mockito
• Spring Boot context loading tests

Code coverage is measured using JaCoCo.

---

Running the Application

1. Clone the repository

git clone https://github.com/addee1/AdamAI-chatbot.git

---

2. Configure Environment Variable

The application uses environment variables to securely manage the OpenRouter API key.

Inside:

src/main/resources/application.properties

the following property is already configured:

ai.api.key=${OPENROUTER_API_KEY:test-key}

Add your API key in IntelliJ IDEA

1. Open:

Run → Edit Configurations

2. Select the Spring Boot application

3. Add the following environment variable:

OPENROUTER_API_KEY=your_api_key_here

4. Save and run the application

---

3. Run the application

mvn spring-boot:run

---

Frontend

The application includes a custom frontend chat interface where users can:

• send messages
• choose AI personalities
• maintain conversation memory
• interact with the AI in real-time

---

Author

Built and designed by Adam Ottosson

Portfolio: https://adamottosson.se