Increasingly, modern software systems mediate interactions between autonomous parties, e.g., in health, e-business, and finance. Autonomy motivates decentralized, loosely-coupled software systems that support flexible interactions between the parties. However, realizing systems with these properties is one of the biggest software engineering challenges today. In particular, popular approaches for modeling interactions (e.g., UML and state machines) lack high-level abstractions; cannot support flexible interactions; and go against basic ideas in distributed systems.

Interaction-Oriented Programming (IOP) is the only game in town that rises to the challenge. Its bedrock is declarative abstractions that capture the business meaning of interactions via notions such as commitments between the parties, e.g., to deliver goods upon payment, supported by decentralized, flexible interaction protocols. IOP is supported by a publicly available tool suite, including compilers, verifiers, and middleware-supported programming models, which will be demonstrated in the workshop.

IOP closes the abstraction gap between stakeholders and programmers and takes the notion of letting programmers focus on the business logic to a whole new level. Multiagent systems are trending once again due to rise of "agentic." The workshop will equip you with the knowledge to engineer AI agents that act flexibly in the context of a multiagent system.