Telehealth Foundry
From patient engagement to payable order.
ProvenProven · Operating
The hard part of remote diagnostics is not the test. It is everything around it: qualifying the patient, documenting the encounter, completing the requisition and establishing medical necessity before an order ever reaches a laboratory. That chain has been built, operated and measured.
What is the true cost of turning a patient into a clean, documented, payable order — and how far can AI take it down?
Genetics · Diagnostics · Remote care
The industry problem
The order is where the value leaks.
Demand generation, clinical encounter, documentation and reimbursement are usually four different businesses stitched together by email, portals and rework. Every seam costs money: leads that evaporate, encounters that are charted late, requisitions that arrive incomplete, claims that bounce on medical necessity.
Four conditions have only recently aligned: compliance is strict enough that doing it properly is a genuine advantage, AI is now capable of carrying the documentation and integrity work, policy has moved on what is reimbursable, and patients expect remote care by default.
The economics
Where value is created, destroyed and trapped.
The economics of a telehealth-fed diagnostics business are decided long before fulfilment. Value is created when a patient is genuinely activated and clinically served; it is destroyed in rework, chasing and denials; and it is trapped in the gap between what is billed and what is actually paid.
Where can AI fundamentally change the economics?
Created
Activated patients, documented encounters, complete requisitions, defensible medical necessity.
Destroyed
Evaporated demand, late charting, incomplete orders, denied and re-worked claims.
Trapped
The gap between billed and collected — and demand that swings between idle and overloaded.
The Foundry thesis
ThesisThe chain should be one system, not four businesses.
If patient engagement, the clinical encounter, documentation and revenue integrity run as a single AI-native operating system with clinicians accountable at the points that matter, then an order can arrive complete the first time — and the cost of each incremental order falls rather than rises with volume.
The system
Inside the operating system.
Four agent-run capabilities operate ahead of fulfilment. Clinicians remain responsible for clinical judgment; the agents carry the labour, the recall and the consistency.
Patient engagement
The order comes from a patient who actually showed up.
Calls, texts, schedules and answers, so demand is activated and seen rather than lost between touchpoints.
Ambient documentation
The order is backed by a real, recorded clinical encounter.
The visit is charted as it happens, under the clinician's review, instead of reconstructed afterwards.
Requisition completion
The requisition arrives complete the first time.
Supporting documentation is generated from the finished chart — nothing missing, nothing to chase.
Revenue integrity
The claim gets paid rather than bounced.
Medical necessity is evidenced and payer rules are tracked as they change, so submissions stay current.
What we’ve built & what we’re building
Clear about what exists today.
Proven results, work underway, capability in development, researched thesis, and where we want partners — labelled, not blurred.
Proven
- A genetics and telehealth ecosystem built and operated at scale: ~$1B ecosystem transactions, ~250,000 patients
- Commercial value chain redesigned to roughly 4× annual revenue capture within a year
- Full-service model in which clinical, laboratory, marketing and operations ran as one flow
Current
- Operating the engagement → encounter → documentation → integrity chain with live partners
- Agent-assisted documentation and revenue integrity in day-to-day use
Building
- Broader payer rule coverage and automated policy monitoring
- Onboarding path for additional diagnostics and device supply partners
Invitation
- Laboratories, diagnostics and device suppliers who want qualified, documented, payable order flow
- Clinician groups who want the administrative load carried without losing clinical control
The people
Where humans stay responsible.
- Clinicians own clinical judgment, the encounter and what is signed
- Compliance leadership owns defensibility and audit posture
- Commercial leadership owns which supply relationships are worth building
- GKIM and Ian own the model of the chain and where it is redesigned next
The AI
Where machines change the economics.
- Carrying engagement, recall and follow-up at a cost per patient humans cannot match
- Charting and document generation from the encounter as it happens
- Checking every order against current payer and medical-necessity rules before submission
- Smoothing throughput so demand can be staffed and forecast rather than absorbed
Human judgment first. Machine acceleration second.
The opportunity
What could emerge from this Foundry.
- 01
Diagnostics and device businesses that can grow order volume without building the machine behind it
- 02
New remote care pathways that were previously uneconomic to serve
- 03
A defensible, documented order standard that becomes worth more as more participants join
Designed with GEARS™
GKIM’s method for digitizing business ideas and industries. Industry expertise goes in, GEARS structures the economic system, and the Foundry turns the resulting knowledge into businesses.
Participate
Who should get in touch.
This Foundry is operating, so conversations here are practical: what you supply, what the flow is worth, and how quickly you could be in it.
We go deep with a small number of businesses and Foundry partners at a time. New relationships begin when we believe there is something consequential to build.
Who should get in touch
- Laboratories and diagnostics operators
- Device and service suppliers
- Clinician groups and telehealth operators
- Investors in regulated diagnostics