Onboard It Like a Hire: The 30-60-90 Plan for Your First AI Agent

Practices are taking on their first non-human workers right now, and most are making one of two mistakes. Either the new tool gets chart access in week one because the demo was impressive, and eventually writes something nobody checked. Or it gets so little access, out of entirely reasonable fear, that it produces nothing and quietly proves to every clinician in the building that “AI does not work here.”

Both mistakes have the same cause: treating the tool like software, when the situation is closer to a hire. And for hires, a clinic already knows exactly what to do, because it has onboarded people into roles with patient contact before.

You already have the playbook

Think about what happens when a promising stranger joins your team. They get a job description, not “help out everywhere.” They get a manager who reviews their work. They get the access they need for their job and nothing more. They spend a probation period being watched more closely than feels efficient. And they earn autonomy by track record, not by charisma on day one.

Every one of those controls maps directly onto an AI tool, and none of them need inventing. A digital worker is new; supervising one is the same discipline a practice already applies to a new locum.

The 30-60-90 shape gives it a rhythm:

The probation ladder

1. 1

   Days 1-30: shadow and draft Start here

   Read-only chart access. It drafts, a clinician signs. Every output reviewed.

2. 2

   Days 31-60: act with approval

   Routine administrative cases queue for a human tap. Anything clinical still escalates.

3. 3

   Days 61-90: own the routine

   Administrative work inside guardrails. Monthly sampling replaces per-item review.

4. 4

   After 90: reviewed, not watched

   A standing owner, a monthly review packet, and a tested revoke switch.

The gates are evidence, not anniversaries

The dates in a 30-60-90 are the cadence for asking the question, not the answer. An agent gets promoted when its file says so, and its file is the eval results: the same fixed set of real cases, scored every week, plus the live record of how often a human had to correct it. That is a performance review a machine can actually sit for, and it is far more honest than a gut feeling that “it seems fine lately.”

This is where the hire metaphor pays for itself, because the promotion criteria almost write themselves. Would you widen a new receptionist’s authority if you were still correcting half their bookings? Then do not move a tool to act-with-approval while clinicians are still editing its drafts. The autonomy rungs themselves are the product-design side of this story; the onboarding plan is the management side: who reviews, on what cadence, and what the packet must show before the next rung.

The two clauses teams forget

The supervisor clause. Every tool needs a named human owner, chosen before launch, and for anything clinical that owner is a clinician. Not a committee, not “the practice”: a person whose job includes reading the monthly eval report and whose name is on the decision to promote or stop it. Tools without owners drift, in the same silent way an unwatched model does: nothing breaks loudly, quality just erodes until somebody reads back through a chart. Ownerless agents are also how sanctioned AI decays into the same ungoverned mess as shadow AI, just with a logo on it.

The offboarding clause. Before the tool touches a chart, know how you take its access away, and test that it works. One switch, not a scavenger hunt through vendor portals. Most practices plan the rollout; almost nobody writes the shutdown plan, and the difference shows up at the worst possible moment. A tool you can revoke in one step is a tool you can afford to trust with more, and it is what Ontario’s privacy commissioner had in mind by recommending clinics keep the authority to pause or decommission an AI deployment.

Give it the onboarding you would give a promising stranger, because that is exactly what it is.

What to do Monday

Pick your first or riskiest tool and write its staff file: one paragraph of job description, the named owner, the current rung, and the promotion criteria in checkable terms. It fits on one page and takes an hour. Then put a recurring 30-minute review in the owner’s calendar, on a fixed day of the month. That single page is the difference between a workflow that climbs safely and the two failure modes every practice is currently choosing between: the tool nobody trusted, and the one everybody trusted too soon.

---

Bringing a first digital worker into your practice? Book a free consult and we will write the job description, the probation plan, and the promotion gates for one real workflow together. See more on Staff Adoption.