Agent Identification: A Domain-Driven Architecture Approach
Shivakumar R Goniwada · Apress eBooks · 2026
Agent identification is not the act of labeling every automation, workflow, or microservices as an agent. It is the disciplined process of discovering the right agent boundaries from the domain itself. It examines the business capabilities, decisions, events, exceptions, policies, and human judgment points to determine where an agent can add meaningful intelligence, autonomy, and adaptability. In this chapter, I will provide the agent identification through the lens of domain-driven architecture not as theory but as a practical compass for building systems that behave like they would in a real organization. Domain-driven design (DDD) taught us to understand complex domains by carving them along their natural joints, separating entities that maintain identity, value objects that express meaning, aggregates that protect business consistency, and bounded context that preserve the language and intent of each part of the domain. Agentic AI introduces something new such as cognitive actors that reason, decide, and collaborate. The challenge is to translate the static structures of the domain into the dynamic anatomy of an agentic system without breaking integrity, without blurring ownership, and without drifting into agent washing.