— Door 04 · For the founders
One specialist.
Hundreds of clinics.
The honest question behind a fast-growing, bootstrapped company: how does one onboarding person give every clinic a real, human setup without becoming the bottleneck? The answer isn't "work faster." It's an AI-assisted workflow that absorbs the routine half — drafting setup scripts and prompt libraries, prepping call briefs, generating follow-ups, watching health signals — so the hours go to the half only a human can do: the confidence call, the empathy, and the judgement on who needs me today. You sell an AI that frees up reception. I'd onboard with an AI that frees up me — same idea, applied to the role.
— The split
What the AI does · what stays human
This is the whole pitch. AI is leverage on the mechanical work. It is never the relationship, and never the call with a nervous clinic owner.
AI accelerates · the routine half
Setup at volume
- Draft per-clinic scripts & prompts. Generate the call-flow config from the clinic's intake answers — I review and tune, never ship raw.
- Call-prep briefs. A one-page brief before every onboarding call — clinic context, services, likely edge cases — auto-assembled.
- Follow-up sequences. The Day 7, Day 14, Day 30 touches drafted in my voice, ready to personalise and send.
- Plain-English explainers. Generate the "explain it to a 65-year-old" answers per question, on demand.
- Health-signal watch. Surface the amber/red clinics each morning so I call the right ones first.
- Knowledge base. Turn every solved issue into a reusable doc — the playbook builds itself as I go.
Human keeps · the half that matters
Trust & judgement
- The confidence call. Looking a nervous clinic owner in the eye (on video) and making them feel safe. Never automatable.
- Reading the room. Hearing the hesitation a metric misses, knowing when to slow down.
- The save. The real conversation when a clinic is about to churn — empathy, not a sequence.
- Judgement on priority. Who genuinely needs me today vs who's fine — the human call on the AI's flags.
- Founder feedback loop. Carrying what I hear from clinics back to the CEO and CTO to improve the product.
- Tone & care. Every clinic feeling like a person set them up, because one did.
— Where the hours go
A full-time week, honestly budgeted
Roughly how a week splits once the AI workflow is running. AI compresses prep and admin, which buys back hours for the calls and the at-risk clinics — the work that actually protects revenue.
Onboarding & confidence callsthe human core · video + phone
~15h
Setup & configAI-drafted · I tune per clinic
~9h
At-risk & save callsthe retention cadence
~8h
Follow-ups & supportAI-drafted · I personalise
~5h
Playbook & founder loopdocs · product feedback
~3h
The trade: without an AI workflow, setup and admin would eat the week and the calls would get rushed. Compressing the routine half is what lets the human half — the calls that retain clinics — get the most hours.
— The stack
Real tools, already running
Nothing speculative — this is the stack I work in daily, pointed at onboarding.
— The honest line
I'm not pitching a bot that replaces the onboarding person. AI handles the mechanical half so the human half gets all of me. A nervous clinic owner handing over their phones can't be reassured by automation — and I wouldn't try. The confidence call, the save, the care: that's the job, and it stays human.
What the AI workflow buys Aeva is simple: one specialist who can give hundreds of clinics a genuinely human onboarding — at the pace a bootstrapped, fast-growing company actually needs.
Honest note: Hour splits and percentages are illustrative planning assumptions for a typical full-time week, not a fixed commitment — the real mix flexes with clinic volume. The AI tools listed are ones I genuinely use daily; Cliniko and the Aeva product are an honest first-weeks ramp, not day-one expertise. This is concept work for my application, not Aeva's official onboarding system. See Door 01 for the full fit map.