Day One

The question

Onboarding a clinic usually means three or four discovery calls, spread over about two weeks because of scheduling. Half of what was said on the first call is lost by the third. The time isn't in the minutes. It's in the waiting between calls.

What I built

An internal platform for onboarding clinics and pharmacies. It runs the discovery call, captures as much as possible while it happens, and turns that into workflows, diagrams, strategy and a plan. Contracts, execution and monitoring sit in the same place.

How it works

1. A structured discovery call captures how the clinic really works, and what it wants to achieve first.
2. The answers become workflows and diagrams.
3. Those become a strategy and an onboarding plan.
4. As soon as the call ends, the clinic's details go straight out for checking, sign-off and confirmation.

AI does the routine, manual work: drafting diagrams, workflows and plans. Anything sensitive is filtered with traditional methods first. Local models and frontier cloud models share the rest.

What happened

We have onboarded one clinic with it so far. The first discovery took under 90 minutes, and the calls after it moved the work forward instead of repeating discovery. The same ground would have taken a week or more, and three or four calls.

The calls are only the start. Diagrams, strategy and planning follow. But the first call now captures enough to move.

It is used internally for now. We will use it with 10 to 20 more clinics and validate it before it goes public.

What I learned

Building it taught me how to run a discovery session: where the time goes, and which questions really matter. The goal is to decide what we build first and what outcome it should bring, then shape the questions and flow around that.

There's no point mapping everything up front. Once phase one starts, things change. Understand the pain points and what the business wants, get that in one short session, then plan, build, validate and let phase two grow from there.