Surface CKD Risk Before Your Patients Feel It
The routine panels your clinic already draws become a prioritized queue for confirmatory kidney evaluation — which patients to move to the front of the line for the eGFR and urine albumin testing KDIGO already recommends. No additional screening order is needed to generate the prioritization; confirmatory testing follows a flag.
The Gap in the Record
The record marks an unmeasured patient as unremarkable
One patient, one draw, read two ways. Nothing on either side was ordered specially — the difference is which values get read together.
Today
What the record shows today
Each result read on its own, and the one the record never received.
Routine results — unremarkable, one at a time
Urine albumin (uACR) — not ordered
No flag raised
Nobody decided this patient does not have CKD. Nobody looked.
Read together
The same record, read together
The same values already on file, entered or uploaded into NephroSense AI.
Prioritize — confirmatory kidney evaluation
Suggested next step: eGFR + urine albumin (uACR)
For Health Systems
What this is measured on
A flag is only worth what happens after it. These are the outcomes a deployment is evaluated against.
Confirmatory-testing completion
Of those flagged, how many received the eGFR and urine albumin testing KDIGO recommends.
Confirmed-risk yield
How many flagged patients were confirmed on follow-up — whether the flag is worth acting on.
Referral completion
Whether patients needing nephrology or care-management follow-up actually reached it.
Workflow burden
Review time per flagged patient, and whether the flag arrives inside an existing workflow.
Solutions
Two products, one platform
VANTAGE runs today, beside the chart. CONDUIT puts the same output inside it, and is in development. Different buyers, and each carries the stage it has reached.
VANTAGE
Available todayCKD risk triage from the panel you already drew
The primary workflow reads a routine adult workup — blood work and vitals already recorded in ordinary care — and returns a CKD risk score with KDIGO-aligned clinical considerations in under ten seconds. No additional screening order, no new hardware. The output is a prioritization: which patients should move to the front of the queue for the confirmatory eGFR and urine albumin testing KDIGO already recommends.
Instant risk assessment
Enter adult patient biomarkers or upload a lab report; a CKD risk score returns in under ten seconds.
Clinical considerations
KDIGO-aligned considerations sorted by urgency — high, moderate, routine — each with a next step.
Confidence signal
Every prediction carries an explicit confidence signal, so the basis for the score is visible rather than assumed.
What-if simulator
Adjust biomarker values and see how the risk score responds.
Reports and exports
A PDF with the risk score and considerations for the patient record, plus FHIR R4-shaped output for exchange with existing systems.
Follow-up planning
Prioritize confirmatory testing, referral urgency and longitudinal monitoring under physician oversight once a patient is flagged.
Generated Report
CKD Risk Assessment Report
Patient ID: NS-2024-0847 · Generated Mar 15, 2026
Illustrative — synthetic data. Not a real patient or a performance claim.
What-If Simulator
Model score response
Current input state
Alternative input state
Illustrative — synthetic data. Not a real patient or a performance claim.
Where this stops: VANTAGE is Clinical Decision Support, not a diagnostic. It does not replace urine testing, eGFR or any other test, and it does not diagnose CKD. Confirmatory kidney testing is ordered by the clinician after a patient is flagged, and every clinical decision remains the provider's, based on a complete patient evaluation.
Clinicians, primary care groups, FQHCs
CONDUIT
Designed, in buildThe same output, inside the chart instead of beside it
VANTAGE runs alongside the chart; CONDUIT would run inside it. It is the embedded layer, returning structured risk into the record and following what happens after a patient is flagged. That integration work is underway now.
Automatic lab ingestion
Reading results from the EHR as they post, so a patient is scored on the panel already in the record rather than on values re-entered by hand.
Embedded order sets
Generating referrals, confirmatory tests and treatment reminders the moment a patient is flagged, configurable to a group's standard of care so the clinician reviews and signs.
Referral routing and closure
Following a flagged patient through to whether the referral and confirmatory test completed — the outcomes a deployment is evaluated on.
Order Set
Auto-GeneratedPatient flagged: High CKD Risk
Referrals
Confirmatory Tests
Treatment
Configurable to your standard of care
Illustrative — synthetic data. Not a real patient or a performance claim.
What has to be true first: A first site. The engineering is in progress; what it needs alongside that is an institution willing to co-design the workflow placement and do the interface and governance work any EHR integration requires there.
Scope: CONDUIT changes where the output lands, not what the model does or what it is permitted to claim. Until it ships, VANTAGE reaches clinicians through the standalone workflow and the FHIR R4-shaped output it already emits.
CMIO, clinical informatics, IT
Evidence
Where the evidence stands
Stated plainly, because you should not have to ask for it.
Retrospective feasibility demonstrated on a US adult clinical cohort. Independent external validation is in progress and is not yet complete.
What this is, regulatorily
NephroSense AI is a Clinical Decision Support tool for use by licensed healthcare professionals with physician oversight. It does not provide medical diagnoses and does not replace any laboratory or diagnostic tests. By integrating a patient's existing clinical data faster, it helps clinicians prioritize who needs confirmatory testing and follow-up sooner. All clinical decisions must be made by qualified healthcare providers based on a complete patient evaluation.
One Platform
One platform, whichever question you arrived with
Health systems and trial sponsors ask different questions. Each application requires its own model and its own validation, but all of them are built on one platform — the same routine-panel inputs, the same deployment architecture, the same IP. These are its properties in both settings.
Routine inputs, already on file
The model reads values already present in a routine adult workup — drawn and paid for as part of ordinary care. No new assay, no new visit, no new hardware.
Model-to-data by architecture
Institutional deployments are built so inference runs inside your environment as a query-only container — patient-level data does not leave it, and the model is not transferred to you as weights.
Patent-pending technology. Provisional patent filed March 2026.
Getting Started
Three steps, and you can stop after any of them
Nothing here requires a commitment to the next thing. The first one does not require talking to us at all.
Run it yourself
Open the demo application and put your own values through it. Nothing to install, no conversation required.
What it settlesWhether the output is something your clinicians would act on.
Retrospective look-back
We run the model against a de-identified extract of your own patients and report what it would have surfaced, against what actually happened.
What it settlesWhether the signal is there in your population, before any deployment.
Scoped pilot
A defined cohort, a defined workflow placement, and the outcome measures above agreed in advance.
What it settlesWhether it changes confirmatory-testing and referral completion in practice.
Before procurement asks
- Deployment
- Output is built on HL7 FHIR R4 for clinical data exchange. Embedded EHR deployment is in development.
- Agreements
- BAA and data-use agreements executed per engagement, with the scope of data covered agreed in writing before any extract moves.
FAQ
Frequently Asked Questions
Running a kidney or cardio-renal trial? The same platform surfaces a candidate population for confirmatory screening — patients with no CKD diagnosis code and no urine albumin result on file.
For Biopharma →See what it surfaces in your own population
A retrospective look-back reports what the model would have surfaced, against what actually happened.