Most organisation charts begin with job titles. An Agentic team begins somewhere else: with the outcomes the business needs, and a clear decision about who or what is best placed to deliver each one.
At V1, our own A-Suite was built before human counterparts were hired for the matching roles. That order matters. It meant each role was defined carefully first, and the humans who joined had something clear to step into. Here is the process we would recommend to any founder starting from a blank page.
Step one: list outcomes, not positions
Write down what must happen every week for the business to work. Clients served, invoices sent, content published, reports produced, decisions made. Roles come later. Starting with titles locks you into yesterday's structure, while starting with outcomes shows you what the work really is.
Step two: sort by who should own it
For each outcome, decide whether a human or an Agentic Citizen should own it. A helpful rule is that humans own judgment, relationships and final accountability, while Agentic Citizens own execution, preparation, monitoring and memory.
- A human should own decisions with legal, financial or reputational consequence.
- A human should own the relationships that depend on trust.
- An Agentic Citizen can own recurring, well defined work where consistency matters.
- Specialists can take the detailed tasks inside that work.
Step three: write a Role Charter
For every Agentic Citizen, write what it is responsible for, what it may decide alone, what it must escalate and who reviews its work. A charter is short, and it is the single most useful document in the design.
Here is an illustrative charter for an Agentic client onboarding role:
- Responsible for: preparing welcome material, collecting documents, scheduling the kickoff and tracking open items.
- Decides alone: reminder timing, document checklists and routine scheduling.
- Escalates: pricing changes, scope disputes and any client concern about quality.
- Reviewed by: the human head of client success, weekly.
Step four: pair each human role with an Agentic counterpart
This is the heart of the SynAgentic model. Each human role has an Agentic counterpart that holds context, prepares the ground and carries execution. The human spends time on judgment and relationships, and the counterpart keeps the machinery moving.
Step five: set a review rhythm
Agentic work still needs oversight. Agree when a human reviews output, what gets sampled and what triggers an escalation. A short weekly review catches drift early, and it builds the trust that lets you extend responsibility over time.
Step six: keep the knowledge yours
The business should own its record of how it thinks and decides. Design so that your knowledge can live in storage you control and is not locked inside a single platform. The model processes the thought, and the organisation owns the record.
Common mistakes
- Giving an Agentic Citizen a vague mission and expecting it to work out the rest.
- Skipping the review rhythm because the early results look good.
- Handing over decisions of real consequence before trust has been earned.
- Redesigning the whole organisation at once.
Start small
Pick one function, build one pairing, review it for a month and learn. The second pairing will be faster and better than the first, because the charter, the review rhythm and the habits are already in place.
If you would like help designing your first pairing, the next step is a Discovery conversation.
