Designing how people and agents work together.

Start with who the work serves, what it should accomplish and how it happens today. Follow the journey across people and systems to understand where context gets lost, effort repeats and decisions depend on someone’s expertise. Question whether each step still serves the outcome as agents change what becomes possible.

Identify the facts, principles and rules that should guide each task. Define when they apply, which sources carry authority and what happens when guidance is missing or conflicting. Some requirements become instructions or checks. Others remain decisions for the people responsible.

Establish what an agent can do, what it needs to carry forward and when someone should intervene. Design the handoffs so the next person or agent can understand the work, its status and the decisions behind it.

People need to direct work, inspect results, correct mistakes and continue from earlier attempts. Those experiences develop alongside the underlying systems, with integration limits, permissions and reliability informing what is practical.

I write to form a point of view, prototype to explore a possibility and build to expose its requirements. Testing can reveal unnecessary steps, missing knowledge or assumptions that change the opportunity itself.

The aim is to expand what a team can deliver while carrying its intent, standards and expertise into the result. What we learn should improve the next version and make clear what is worth building, what it requires and what still needs to be tested.