Agent-oriented software engineering (AOSE) is a promising approach for developing complex, distributed systems. This article argues that modeling software systems as interacting, autonomous agents offers significant advantages over traditional methods. While AOSE is not a silver bullet, it has shown potential in various applications. The complexity of industrial-strength software systems is inherent, and managing it requires techniques that allow for decomposition, abstraction, and organization. AOSE aligns well with these principles by decomposing systems into autonomous agents, which can interact and collaborate to achieve their objectives. AOSE is particularly well-suited for complex systems due to its ability to model interactions at a knowledge level, enabling flexible and adaptive behavior. Agents can negotiate, coordinate, and manage dependencies in a dynamic environment. The approach also supports the modeling of organizational relationships, which is essential for managing the changing nature of complex systems. AOSE provides explicit structures for representing and managing these relationships, making it suitable for applications involving multiple stakeholders. AOSE has been successfully applied in various domains, including telecommunications, where it has been used to provision virtual private networks. In such applications, agents represent different entities, such as end users, service providers, and network providers, and interact to achieve common goals. The approach allows for the dynamic negotiation of resources and services, demonstrating its effectiveness in managing complex, distributed systems. The article concludes that AOSE is a valuable paradigm for software engineering, offering a natural way to model complex systems and manage their interactions. While it is not a complete solution, it provides a flexible and adaptive framework that is well-suited to the challenges of developing complex software systems.Agent-oriented software engineering (AOSE) is a promising approach for developing complex, distributed systems. This article argues that modeling software systems as interacting, autonomous agents offers significant advantages over traditional methods. While AOSE is not a silver bullet, it has shown potential in various applications. The complexity of industrial-strength software systems is inherent, and managing it requires techniques that allow for decomposition, abstraction, and organization. AOSE aligns well with these principles by decomposing systems into autonomous agents, which can interact and collaborate to achieve their objectives. AOSE is particularly well-suited for complex systems due to its ability to model interactions at a knowledge level, enabling flexible and adaptive behavior. Agents can negotiate, coordinate, and manage dependencies in a dynamic environment. The approach also supports the modeling of organizational relationships, which is essential for managing the changing nature of complex systems. AOSE provides explicit structures for representing and managing these relationships, making it suitable for applications involving multiple stakeholders. AOSE has been successfully applied in various domains, including telecommunications, where it has been used to provision virtual private networks. In such applications, agents represent different entities, such as end users, service providers, and network providers, and interact to achieve common goals. The approach allows for the dynamic negotiation of resources and services, demonstrating its effectiveness in managing complex, distributed systems. The article concludes that AOSE is a valuable paradigm for software engineering, offering a natural way to model complex systems and manage their interactions. While it is not a complete solution, it provides a flexible and adaptive framework that is well-suited to the challenges of developing complex software systems.