Skip to main content
HealthEquation
Case Study · Technology

Building IntelliHealth™, and keeping it assistive.

How we built a clinical decision engine that drafts, ranks, and flags — while a pharmacist keeps final say on every meaningful decision.

15 min read · Field note

IntelliHealth™ is our clinical decision-support system. It analyzes the patient record, proposes protocol options, and flags issues worth a second look — and it routes every significant call to a pharmacist for sign-off.

The design rule
The system is assistive, not autonomous. It can draft and surface; it does not decide. A licensed clinician reviews and approves anything that touches a patient.

The problem

Personalized care is hard to scale. As patient volume grows, the time per patient shrinks, and quality is usually what gives. We wanted a system that protected the clinician’s attention — handling the gathering and the first pass so the human could spend their judgment where it matters.

How it’s built

Three layers do the work. None of them act on their own; each one prepares the ground for a person to make the call.

Decision engine

Models that surface and rank protocol options for review — never auto-prescribe.

Data layer

A patient record that assembles the context a pharmacist needs in one place.

Integration

Connections into the systems clinicians already use, so the work fits the workflow.

What it handles

120K+
Patients / mo at Nimbus
Draft
+ flag, not decide
100%
Clinician sign-off

What we learned

The hardest part was never the modeling. It was drawing a clear line between what the system suggests and what a person approves — and holding that line as volume grew. The value of the tool turned out to be the time it gave back to clinicians, not the decisions it could make on its own.

A fuller technical write-up — architecture choices, the trade-offs, and the failure modes we designed around — is in progress and will land here.

The Equation

The engine handles the pattern. You handle the patient — IntelliHealth never gets ahead of that.

Want the architecture details?

The free AI Literacy course covers the ideas behind keeping AI assistive in clinical work. If you’d like to talk implementation, the door’s open.