Agents are becoming increasingly common as AI capabilities make them easier to build and deploy. As a result, a familiar engineering question has returned: not whether we can build agents, but whether we should.

This session examines the practical considerations involved in designing systems that incorporate agents. We will explore different types of agents, the roles they can play within software systems, and the situations in which agent-based approaches are most appropriate. The discussion focuses on architectural decision-making rather than implementation details, helping attendees evaluate where agents provide value and where alternative approaches may be more suitable.

The talk also considers how software systems can be designed and evolved as different types of agents are introduced over time. Through a discussion of architectural patterns and trade-offs, attendees will gain a clearer understanding of how agents fit into broader system design and how to make informed decisions about their use.

What You Will Learn:

• The different types of agents and the roles they can play within software systems
• How to evaluate where agent-based approaches are appropriate and where they may not be
• Architectural considerations for designing and evolving systems that incorporate agents

Who Should Attend:

• Software engineers
• Software architects
• Technical leads
• Engineering managers
• AI engineers
• Platform engineers
• Teams evaluating or adopting agent-based systems