Managing AI Agents: Roles, Memory, Context and Handoffs

Starting with an AI agent is easy. Give it a task, some context and access to the right tools, and it can already do useful work.

But over time, things get messy. An agent may lose important context, remember something it should have forgotten, work with outdated information or continue a task based on an assumption that is no longer true. If several agents are involved, they may duplicate work, miss handovers or act on different versions of the same information.

Does that sound familiar? Yes, it is a bit like managing a badly organised team.

In this online session, we will discuss how to manage AI agents in your communications team. What context, memory and state actually mean, where information should live, how agents hand work over to each other and to people, and how to keep everyone working from the same version of reality.