AI employees built for nephrology groups. They prepare each patient's chart before the visit, take referrals from fax to first appointment, and answer patient calls around the clock. Everything is written back to your EHR, and your clinicians keep the judgment.
Modeled for a four-physician nephrology group. How we model it
Example
Chart ready before the visit
Written to your EHR before the patient arrives.
What nephrology groups ask for first
The two jobs that take the most staff time in a nephrology clinic, done by AI employees and written to your EHR.
Chart preparation Available today
Today Staff pull labs, notes and coverage by hand the day before, or the morning of.
Before each visit
Automatic referral management Beta
Today Faxes sit in a tray and get keyed by hand, and missing records turn up the week of the visit.
When a referral arrives
The Nephrology OS, end to end
Live workflows first, then the ones in beta. Each agent takes the repetitive part of one job and hands anything clinical or unusual to your team.
| AI employee | Today | With EigenH AI | Handled by AI | Revenue at risk / yr |
|---|---|---|---|---|
| Scheduler Available today | Today Calls missed mid-morning and after hours. | With EigenH AI Every call answered 24/7; visits booked inside your rules. | Handled by AI ~80% | Revenue at risk / yr ~$101,700 |
| Recaller Available today | Today Staff chase the same patients by phone. | With EigenH AI Outreach on a cadence; recurring visits rebooked automatically. | Handled by AI ~75% | Revenue at risk / yr ~$33,900 |
| Verifier Available today | Today Coverage verified once at intake, then rechecked by hand as plans change. | With EigenH AI Eligibility checked before each visit; changes and gaps sent to staff. | Handled by AI ~50% | Revenue at risk / yr ~$13,600 |
| Referrer Beta | Today Faxes keyed by hand; missing records found late. | With EigenH AI Referrals read on arrival; missing records requested the same day. | Handled by AI ~70% | Revenue at risk / yr ~$39,100 |
| Authorizer Beta | Today Every packet built from scratch. | With EigenH AI Packets assembled for staff; expiring approvals flagged early. | Handled by AI ~50% | Revenue at risk / yr ~$15,600 |
| Nourisher Beta | Today Dietitian referrals wait in a queue. | With EigenH AI Dietitian visits scheduled; diet questions routed to your care team. | Handled by AI ~60% | Revenue at risk / yr ~$14,100 |
Estimates for an illustrative four-physician nephrology group, not a guarantee of results. Actual results depend on your volumes and are confirmed before go-live.
Inputs: $135.61 per established-patient visit (CPT 99214, CY 2026 Medicare national non-facility rate, CMS); 250 working days. Scheduler: 40 inbound calls a day, 3 booking calls missed a day. Recaller: 5 recurring visits slipping a week. Verifier: 60 visits a day; 2 visits a week lost to a coverage problem found after the visit (modeled assumption). Referrer: 60 referrals a month, 10% never seen, 4 visits each in year one; published research finds about half of specialty referrals are never completed (Weiner et al.), so 10% is conservative. Authorizer: 2 lapsed authorizations a month at about $650 each (modeled assumption). Nourisher: 8 unscheduled dietitian visits a month at $146.96 each (four 15-minute units of CPT 97802, CMS). Handled by AI shares, the 50–60% and the staff hours are our estimates. Revenue at risk is revenue exposed to these gaps today.
Sources: CMS 2026 fee schedule Weiner et al. 2010
For practice owners
Revenue exposed today to missed calls, slipped visits, lost referrals, eligibility denials and lapsed authorizations, by AI employee.
$218,000 revenue at risk protected a year
Bring last month's call, referral and recall volumes to a demo and we rerun the model with your figures. How we model it
Built for nephrology
Primary care often verifies coverage once and books a visit or two a year. Kidney care does neither, because the disease progresses and takes the schedule and the insurance with it.
Cadence
Recurring nephrology visits, lab draws between them, medication changes and the calls each one generates. The Recaller is built for that rhythm.
Coverage
Patients can move from commercial coverage toward Medicare as kidney disease progresses, often holding more than one plan at once. The Verifier checks before each visit, not just once at intake.
Transitions
Monitoring, modality education, vascular access planning, a transplant referral that starts and stalls. Each transition means new scheduling, a new authorization and a new insurance question.
Nutrition
Dietitian visits are part of kidney care, and they slip when nobody books them. The Nourisher keeps them on the calendar and sends every diet question to your care team.
For nephrologists
Kidney care is exactly the setting where an over-confident agent would do damage, so the boundary is a design constraint rather than a disclaimer.
Anything urgent, clinical, or outside the approved workflow routes to staff immediately, with the transcript and the context attached.
Every call ends with intent, outcome, transcript, and next step. The point is a short list of things that genuinely need a person, not a longer dashboard.
Chart prep gathers results; it does not read them. The AI employees do not judge medical necessity, interpret a lab, or weigh in on modality. Those belong to your clinicians.
Where we are today
Live workflows run today. Beta workflows are set up with your team during onboarding.
Each beta workflow is set up with your team during onboarding and refined as it runs.
Every appointment, update and call outcome lands in the system your team already works in, so nobody retypes what an AI employee did.
Read more
Kidney care is a years-long relationship, not an annual visit. Here is what that does to a nephrology practice's phone line, scheduling rules, and follow-up list, and which parts of it an AI front desk can honestly take on.
In kidney care a missed call is rarely a lost new patient. It is a recurring visit that slips, or a referral that never books. Here is a simple model, with every assumption stated, for estimating what unanswered phones cost a nephrology practice.
Most specialties verify coverage once. Kidney care verifies it repeatedly, because the coverage itself moves as the disease progresses. A look at the 30-month coordination period, the real cost of manual verification, and what changes when CMS prior-authorization APIs land.
It is our platform set up for nephrology: AI employees that run the repetitive work around each kidney patient, from automatic referral management and chart preparation to eligibility, prior-auth paperwork, calls and scheduling, recurring recall and dietitian visits. Every outcome is written back to your EHR.
Yes. Before each visit it pulls recent labs such as eGFR, creatinine, potassium and UACR into the chart, attaches referral letters and the last visit's notes, checks coverage for the visit date, and flags open items for staff. It gathers results and never interprets them.
The Referrer (beta) reads each faxed or eFaxed referral when it arrives, checks it against your intake requirements, requests missing records from the referring office the same day, and hands the patient to the Scheduler to book the first visit. Status, records and the appointment are written back to your EHR, and the referring office is updated.
An answering service takes calls. The Nephrology OS covers the whole patient journey, and each workflow is built for kidney care: recurring visits on a cadence, referrals that arrive without labs, coverage that changes as the disease progresses, and dietitian visits that are part of treatment. It is set up with your team by people who know nephrology.
Chart preparation, the Scheduler (calls and booking), the Recaller (recall and follow-up) and the Verifier (eligibility) are live. The Referrer, Authorizer and Nourisher are in beta and are configured with your team during onboarding.
They are estimates for an illustrative four-physician nephrology group. Visit values come from the CY 2026 Medicare Physician Fee Schedule and the referral input from published research; the other volumes are modeled assumptions, and we never use vendor data. Every input and source is listed under the workflow table on this page. In a demo we rerun the model with your own volumes, and the numbers are confirmed before go-live.
Yes. Every appointment, update, and call outcome is written back to your EHR, so nobody retypes what an agent did.
Routine requests such as rescheduling are completed within your rules. Anything clinical, including possible symptoms of high potassium, goes straight to your on-call path; the agents do not triage. See after-hours calls in kidney care.
The Recaller sends approved reminders and makes rescheduling easy, and every response is recorded for staff. Our post on cutting dialysis no-shows explains the approach and its limits.
No. The Authorizer (beta) gathers what a request needs and tracks payer status; your team reviews and submits every request.
Walk through one patient end to end, from the referral to the recurring visit, and see what your team reviews at each step.