Once your company is past the "experiment with one agent" phase and into the "we have five working agents and a roadmap for ten more" phase, a question shows up. Should we hire someone to own this. If so, who, and where do they sit on the Accountability Chart.
This is one of the few new hiring decisions AI integration creates inside an EOS® company. Worth treating carefully.
The three shapes the seat takes
There are three credible job titles for the role, with different scopes.
Shape one: Chief of Staff (with AI as part of the scope). The person who runs the Integrator's office. Manages the agent layer alongside calendar, briefings, and team coordination. Reports up to the Integrator. Best for companies with 25 to 75 people where the agent layer is meaningful but not the whole job.
Shape two: AI Implementer (or "AI Operations Lead"). A focused seat that owns the agent layer end-to-end. SOPs. Vendor relationships. Trust ladder management. Quarterly Agent Layer Review. Reports up to the Integrator. Best for companies with 50 to 200 people running 10+ agents.
Shape three: AI Director or Head of AI. A senior seat with leadership-team visibility. Owns AI strategy across product and operations. Reports up to the Visionary or the Integrator depending on whether AI is product-strategic or operations-strategic for the company. Best for companies above 200 people or for companies where AI is in the product itself.
Pick the shape that matches your company's stage. Most EOS® companies reading this post are in the first or second shape's range.
What this seat actually does
A working AI Implementer (using that title generically) does six things.
One, owns the agent layer SOPs. Every agent's job description, scorecard, and SOP is current. The seat is the document custodian.
Two, runs the trust ladder. New agents start in shadow mode. The Implementer decides when they graduate to approve-to-send, then to autonomous within bounds, then to full autonomy. Promotions are recorded with evidence.
Three, manages vendor and tooling. Anthropic vs OpenAI vs Microsoft vs Google. Enterprise contracts. Cost monitoring. New model rollouts. The Implementer holds the relationships and makes the recommendations.
Four, runs the Quarterly Agent Layer Review. Same cadence as the EOS® Quarterly. Walk every agent. Retire what is not working. Plan what to add next quarter.
Five, manages the company-wide preamble file. Same-day refresh after Annual and Quarterly sessions. V/TO™ change notes filed. Agent drift detection.
Six, coaches the human team on hybrid workflows. Most of the team will need to learn how to work with the agent layer. The Implementer runs the enablement. Less than you would think. Some of it is just "here is where the brief is each morning."
What this seat does not do
Three things this seat does not own.
One, V/TO™ authorship. That belongs to the Visionary and the leadership team. The AI Implementer reads it, propagates it, and protects it. Does not write it.
Two, the People Component for the human team. The Integrator and the human-team managers own People. The AI Implementer is responsible for the agent seats only.
Three, the strategic decision of which seats become agent-led. That is a leadership team decision. The AI Implementer can propose. The leadership team chooses.
The line is clean: the AI Implementer is an operations seat, not a strategy seat. Same as a Director of Operations.
What to look for in a candidate
Six traits.
Trait one: practitioner experience with one or more agent platforms. Has built or run agents on Claude Code, OpenAI Assistants, Copilot Studio, or similar. Not just "used ChatGPT." Has actually shipped something that ran on a schedule.
Trait two: written SOP discipline. Can read an SOP and tell you what is missing. Can write one a new hire (human or agent) can execute. This is unusually rare.
Trait three: comfortable saying no. Many requests for new agents are bad ideas. The right Implementer pushes back, surfaces the negative case, and protects the layer from sprawl.
Trait four: numerate on scorecards. Comfortable defining KPIs precisely. Not afraid of definitions discipline. The Scorecard agent layer depends on this.
Trait five: not the smartest person in the room and not trying to be. This is a service seat. The Implementer makes the leadership team's job easier. Ego is a tax.
Trait six: GWC™ on the role. Same EOS® People Analyzer™ test. Do they Get it. Do they Want it. Can they (Capacity to do it). If any leg is yellow or red, do not hire.
Some technical fluency is required but the role is not primarily technical. The Implementer should be able to read a system prompt and discuss it with the technical builder. They do not need to write the prompts themselves.
What to pay
This depends on company stage and geography but a rough read.
Chief of Staff with AI scope at a 25-50 person company. $90K to $140K base in the US. May or may not include equity.
AI Implementer at a 50-200 person company. $130K to $200K base. More likely to include equity.
AI Director at a 200+ person company. $180K to $300K base plus meaningful equity. Senior leadership pay band.
These are not authoritative numbers. Calibrate against your market and your other leadership pay.
Build vs hire
A common question: should we build the agent layer with our existing team or hire someone new.
Build with existing team if: Your Integrator has technical depth and bandwidth, or you have an engineer who genuinely wants to own the operations side. About 30% of small EOS® companies reading this post have this person already.
Hire if: Your existing team does not have the skill or the bandwidth, and the agent layer is going to be 5+ agents by year-end. About 70% of small EOS® companies need this hire.
Do not hire prematurely. A company with one or two agents does not need a dedicated seat. A company with five or more usually does.
What to do in the first 90 days of the hire
If you have just hired an AI Implementer, set them up for success.
Week 1 to 2. Read the V/TO™. Read every existing agent's SOP. Sit in on the L10®. Meet the leadership team. Do not change anything.
Week 3 to 4. Audit the current agent layer. Each agent's scorecard delivery rate over the prior 90 days. Drift signals. SOP gaps. Hand the audit to the Integrator.
Week 5 to 8. Pick one or two agents to sharpen first, based on the audit. Update SOPs. Tighten prompts. Earn the team's confidence.
Week 9 to 12. Start scoping the next agent the leadership team wants. Run it in shadow mode by the end of the quarter.
Quarter 2. Be present for the Quarterly Agent Layer Review with the Visionary and Integrator. Be the one who walks the agent layer with them.
By month six, the seat is the company's institutional memory for the agent layer. By month twelve, the company cannot imagine running without them.
FAQ
Should this seat be remote? Yes, almost always. The work is mostly written. Remote works well.
Should we hire a consultant first and convert to hire later? Reasonable for the buildout phase. Many companies pay an agency or a consultant for the first six months and hire internally once the pattern is in place. Both paths work.
What about contractors or fractional Chief of Staff? Workable for companies that cannot justify a full-time seat. Make sure the contractor has the same accountability lines you would give a full-time hire. Document the same SOPs.
What if the candidate is more technical than the role needs? Fine, often a feature. Over-qualification on technical is rarely the problem. Under-qualification on judgment and SOP discipline is.
EOS®, Entrepreneurial Operating System®, V/TO™, Level 10 Meeting®, L10®, Rocks™, Scorecard, Accountability Chart, People Analyzer™, GWC™, People Component, Visionary, Integrator, Quarterly, and Annual are concepts and trademarks of EOS Worldwide, LLC. This article is an independent practitioner perspective and is not affiliated with or endorsed by EOS Worldwide.