Enterprise Java Patterns: DAO, Repository, Service Layer, DTO, and Unit of Work

DAO and Repository

DAO focuses on abstracting low-level data access operations. Repository is closer to domain-driven design and usually expresses collection-like access for aggregate roots. Many teams use the terms loosely, but understanding the distinction improves design communication.

Service Layer

Service Layer centralizes business use cases and coordinates repositories, validation, transactions, and integration calls. It prevents controllers from becoming too smart and keeps domain workflows in one clear place.

public class OrderService {
    private final OrderRepository orderRepository;

    public OrderService(OrderRepository orderRepository) {
        this.orderRepository = orderRepository;
    }

    public OrderDto placeOrder(CreateOrderRequest request) {
        Order order = new Order(request.customerId(), request.items());
        orderRepository.save(order);
        return new OrderDto(order.getId(), "CREATED");
    }
}

DTO Pattern

DTOs carry data between layers or across the network. They prevent direct exposure of internal entities, make API contracts clearer, and avoid lazy-loading surprises in web layers.

Unit of Work Thinking

Unit of Work tracks a business transaction and coordinates writes as a single logical change set. ORM frameworks often provide this behavior implicitly through persistence contexts and transaction scopes. Understanding the pattern helps explain why frameworks behave the way they do.

Real-World Usage Snapshot

These patterns appear constantly in Spring Boot, Jakarta EE, monoliths, and modular enterprise services. They are not always glamorous, but they are the patterns many Java developers use every day in production code.