
Our AI referral management software reads every inbound referral, matches the patient, and books the appointment. No referral sits in a fax queue for a week. We built it for the administrator who owns new patient volume, referral leakage, and the time from referral received to appointment booked.
MagicDocs, our AI referral coordinator, does the reading and routing. The Referrals AI Agent and Fax AI Agent turn faxed and emailed referrals into scheduled visits inside your EHR. We built this for specialty groups, MSOs, and health systems from 10 to 100+ providers.
Referrals leak because someone has to read, match, and act on each one by hand. One large US health system study makes that gap concrete. Published in the Journal of General Internal Medicine (2018), it found only 34.8% of referral scheduling attempts ended in a documented completed appointment.
Each of those is a referral at risk of never becoming a visit. Our AI referral management software closes that gap.
How it works
We run the whole referral workflow, from the moment a document arrives to the moment the visit is on the schedule. Here are the five steps.
- Ingest documents. Our Fax AI Agent captures every fax, e-fax, portal upload, and scanned record into one queue across all of your locations.
- Extract and match. MagicDocs reads each document, extracts clinical and administrative data at up to 99.5% accuracy, and matches the patient to the chart.
- Classify and complete. We classify the referral, check records readiness, verify insurance, and flag what is missing so nothing stalls between offices.
- Coordinate outreach. Our AI Phone Agent reaches patients and referring offices to gather missing information and confirm scheduling.
- Schedule in the EHR. We write a structured, chart-ready referral into your EHR work queue and book the appointment, in under two minutes per document.
Every extracted field carries a confidence score. Low-confidence items go to staff review instead of getting filed automatically.
A full audit trail records what the AI did and what a person changed. Humans stay on the exceptions by design.
We do not replace clinical or administrative judgment. Staff review flagged exceptions, confirm unmatched patients, and handle payer or policy edge cases. Implementation asks your team for EHR access, a short routing-rule review, and brief onboarding, not a new system to run.
Intelligent referral processing
We process referrals the way an experienced coordinator would, only faster and at scale. Here is what that looks like in practice.
- We read faxes, e-faxes, portal uploads, and scanned records without asking staff to sort them first.
- We extract patient and clinical data from PDFs, images, and handwritten forms at up to 99.5% accuracy.
- We match each patient to the right chart and hold records we cannot match for staff review.
- We chase missing information from referring offices instead of only filing what happens to arrive.
- We track every open referral so none stall between received and booked.
Referral tracking software vs AI referral management
Automated referral tracking software monitors the status of a referral that someone on your team still has to work. AI referral management does the work: it reads the referral, matches the patient, verifies insurance, chases missing information, and books the visit.
| Approach | Who works the referral | What you learn |
|---|---|---|
| Tracking | Your staff read, match, verify, and book each one by hand | That a referral is stuck |
| AI referral management | We do, with staff on flagged exceptions | That the visit is booked |
Faxed referral routing into the EHR
Yes. Our Fax AI Agent captures every fax, e-fax, and scanned record, and MagicDocs splits multi-document faxes, classifies each item, matches the patient, and writes the referral into your EHR work queue. Coordinators never retype a fax by hand.
Fax is still a live exchange channel. According to ASTP/ONC, 40% of US hospitals often used mail or fax to send summary of care records in 2025, and 35% often used it to receive them.
Key features of our AI referral management software
Our AI referral management software combines document understanding, patient matching, and EHR write-back in one platform. These are the capabilities administrators evaluate most. The performance figures below come from Insight Health production data as of 2026.
- Document ingestion across channels: we pull faxes, e-faxes, portal uploads, and scanned records into one work queue across every location.
- Clinical data extraction at up to 99.5% accuracy: we read and structure each document in under two minutes, cutting manual document review by 80%.
- Patient matching to a chart-ready record: we match to the correct chart and hold unmatched records for staff review.
- Confidence scoring and human review: low-confidence fields route to staff, and a full audit trail logs every change.
- Structured EHR write-back: we drop chart-ready referrals into the EHR work queue your team already uses.
- Specialty-aware logic across 80+ specialties: we handle a GI procedure referral differently from a cardiology consult.
Specialty-specific solutions
Referral logic is not one size fits all, so we tune it by specialty. These are three of the most common.
Gastroenterology
- Colonoscopy and procedure classification with an ASC versus hospital determination.
- Recall and backlog clearance so overdue patients get rescheduled.
- Prep and records readiness checks before the visit is booked.
Orthopedics
- Injury classification and priority triage for time-sensitive cases.
- Imaging and records readiness checks before the appointment.
- Referring-office follow-up when a study or note is missing.
Cardiology
- Consult versus procedure routing to the right provider.
- Priority triage for symptoms that should not wait.
- Insurance verification and records checks ahead of the visit.
Outbound specialty referrals
When your practice sends a patient to a specialist, the referral stays open until the consult note comes back. We close that loop in three steps.
We monitor open items on every referral you send, so a missing appointment or report surfaces early.
When the specialist's consult or clinical note returns, MagicDocs reads it, matches it to the patient's chart, and routes it to the right work queue.
If a status or note is still missing, our AI Phone Agent reaches the patient and the specialist's office to follow up.
We do not publish a network-wide outbound referral dashboard, so ask any vendor, us included, to demo outbound status tracking on your own referrals.
CMS tracks this directly. Its MIPS Quality Measure #374, Closing the Referral Loop, measures the percentage of patients with referrals for which the referring clinician receives a report from the specialist.
What AI referral management delivers for your practice
We measure ourselves in the numbers you report to your CFO and your board. In Insight Health deployments as of 2026, MagicFlow has driven a 40% reduction in referral leakage. It has also saved more than 15 hours per care coordinator each week and delivered 3x faster time to a specialist appointment.
In one Insight Health deployment (2026) at a 30-provider gastroenterology group, we cleared a year-long referral and colonoscopy recall backlog. We cut cancellations by more than 16%.
Administrators reduced manual document review by 80%, moved four staff members to higher-value work, and saved $250,000 a year. In that same engagement, clinicians started visits with complete histories, saving 12 to 15 minutes per visit.
For the administrator, that adds up to a short list of outcomes.
- A 40% reduction in referral leakage, keeping more new patient volume in your network.
- 3x faster time from referral received to a booked specialist appointment.
- 15+ hours reclaimed per care coordinator each week for higher-value work.
- 80% less manual document review, which lowers cost per document and headcount as volume grows.
Prior authorization as a supporting workflow
Prior authorization is a supporting workflow that feeds the same referral pipeline, not the headline. It is also where referrals stall today.
The 2024 CAQH Index (calendar 2023 data) puts a manual prior authorization at 24 minutes of provider and staff time by phone, fax, or email. The provider labor cost is $12.88 per manual transaction versus $5.38 fully electronic, or $16.29 versus $5.43 once health plan costs are included.
The American Medical Association 2025 Prior Authorization Physician Survey was fielded in December 2025 among 1,000 physicians. It found practices complete 40 prior authorizations per physician each week. In that same AMA survey, physicians and staff spend 13 hours weekly on prior auth, and 95% report care delays.
Our Prior Auth AI Agent captures the same structured data to keep those requests moving through the pipeline.
CMS-0057-F, the CMS Interoperability and Prior Authorization final rule, requires impacted payers to run a FHIR-based Prior Authorization API beginning January 1, 2027. It covers Medicare Advantage, Medicaid, CHIP, and qualified health plan issuers on the federally facilitated Exchanges, letting practices submit and track authorizations electronically (CMS, 2024). Any prior auth performance figures we share are pilot-stage research targets, not production results.
Developed by doctors
Our platform is built by practicing physicians who are also co-founders, including a neurosurgeon and a cardiologist. They shape the clinical logic behind every referral decision. That is why our AI referral management software handles a GI procedure referral differently from a cardiology consult.
Integrates with your EHR
We integrate with the systems your team already runs: Epic, athenahealth, eClinicalWorks, NextGen, AdvancedMD, DrChrono, Elation, Charm Health, ModMed, and Office Practicum. We write structured, chart-ready referrals directly into your EHR work queue. Coordinators keep working in the system they know, with no new inbox to learn and no swivel-chair between tools.
Structured write-back means the referral arrives as discrete fields, not a PDF someone has to retype. Your staff review exceptions, approve the routine, and move on. See our EHR integrations for the current list of live systems.
How Insight Health compares for referral management
We are HIPAA compliant, which means we safeguard protected health information under US federal privacy and security law. We are also SOC 2 Type II audited. That is an independent review of how we protect data over time.
We run under S.A.F.E., our Supervised Autonomous Framework for Evaluations, with real-time supervision and multi-model verification. Every action lives in a full audit trail, and multi-site groups and MSOs work from one unified queue across locations.
| Capability | Insight Health | ReferralMD | Assort Health | Linear Health | Honeyhealth | Phreesia | Tennr |
|---|---|---|---|---|---|---|---|
| Reads and structures inbound clinical documents | Yes | Referral workflow focus | Phone focus | Yes (OCR-based fax intake) | Yes (fax triage) | Intake focus | Yes |
| Extraction accuracy published | Up to 99.5% | Not published | Not published | Not published | Not published | Not published | Not published |
| Patient matching to a chart-ready record | Yes | Partial | Not published | Not published | Not published | Partial | Partial |
| Structured write-back into the EHR work queue | Yes | Partial | Partial | Yes (real-time) | Yes (HL7/FHIR) | Partial | Partial |
| Automated patient and referring-office outreach | Yes | Partial | Yes (phone) | Yes (voice AI) | Not published | Yes (intake) | Partial |
| Confidence scoring with human review | Yes | Not published | Not published | Not published | Not published | Not published | Not published |
| Full audit trail | Yes | Not published | Not published | Not published | Not published | Not published | Not published |
| HIPAA compliant and SOC 2 Type II | Yes | HIPAA and BAA described, SOC 2 Type II not disclosed | Yes | Not verified in reviewed materials | Not verified in reviewed materials | HIPAA and SOC 2 Type 2 described | Not disclosed |
| Specialty-aware across 80+ specialties | Yes | Not published | Partial | Not published | Not published | Not published | Not published |
| One unified queue across locations | Yes | Partial | Partial | Not published | Not published | Partial | Partial |
Publicly described capabilities as of August 2026. We list what each vendor publishes and mark what is not publicly specified rather than guess.
The best AI referral management software in 2026, reviewed
We are one of the seven vendors below. Each entry uses the same three lines, drawn from what each vendor describes publicly.
1. Insight Health
- Best for: specialty groups, MSOs, and health systems from 10 to 100+ providers where most referrals arrive by fax.
- Core workflow: reads, matches, verifies, chases, and books, with up to 99.5% extraction accuracy in under 2 minutes per document.
- EHR write-back: structured, chart-ready referrals into the EHR work queue across 10 named EHRs.
We are strong at document understanding, with 80% less manual document review in production. Where we differ: the document itself becomes the trigger for action, so work starts the moment a referral lands.
2. ReferralMD
- Best for: health systems and large practices coordinating across provider networks.
- Core workflow: closed-loop tracking, SmartFAX intake, SmartMATCH network matching, and voice AI scheduling.
- EHR write-back: HL7, FHIR, and API integrations with Epic, Cerner, and athenaOne.
ReferralMD is strong at network matching and offers a free entry tier. Where we differ: the document triggers the work rather than creating a task.
3. Assort Health
- Best for: multi-specialty practices that want phone-first coordination.
- Core workflow: voice, SMS, and email referral scheduling, with 29-language support.
- EHR write-back: bidirectional, with 20+ EHR integrations.
Assort Health is strong at voice depth and language coverage. Where we differ: they lead from the call, and we lead from the document.
4. Linear Health
- Best for: practices prioritizing fast implementation on athenaOne.
- Core workflow: inbound fax referral coordination, with voice AI in English and Spanish.
- EHR write-back: real-time, deepest with athenaOne, plus eClinicalWorks, Epic, and Healthie.
Linear Health is strong at speed to go-live, at roughly four weeks. Where we differ: we add document understanding beyond OCR and specialty-aware triage across 80+ specialties.
5. Honeyhealth
- Best for: group practices and MSOs wanting broad back-office automation.
- Core workflow: fax triage, referral intake, and custom agents trained on the customer's workflows.
- EHR write-back: HL7/FHIR, with a claimed 73+ EHR and app integrations.
Honeyhealth is strong at back-office breadth. Where we differ: our clinical document logic is physician-built, and our outreach runs through to a booked visit.
6. Phreesia
- Best for: large health systems bundling referrals with intake and payments.
- Core workflow: referral digitization into one dashboard, tracking, scheduling, and VoiceAI.
- EHR write-back: integrations with Epic, athenaOne, NextGen, eClinicalWorks, MEDITECH, and others.
Phreesia is strong at enterprise footprint. Where we differ: we focus on clinical document understanding rather than digitization and tracking.
7. Tennr
- Best for: DME, infusion, and specialty practices needing eligibility depth.
- Core workflow: faxed referral processing, eligibility and benefits checks, and automated payer phone investigation.
- EHR write-back: public materials describe EHR and eFax integrations and automated data entry.
Tennr is strong at payer-side criteria mapping. Where we differ: we add specialty-aware clinical triage and outreach through to a booked appointment.
Questions about AI referral management
What is the best AI referral management software for clinics?
The best AI referral management software for a specialty group reads inbound documents, matches patients, and books appointments inside your EHR. We built our platform for 10 to 100+ provider groups, MSOs, and health systems, with confidence scoring and humans on the exceptions. It is developed by practicing physicians, so the clinical logic reflects how referrals actually move.
How does AI process incoming clinical referrals?
We ingest each faxed, e-faxed, portal, or scanned referral into one queue. MagicDocs reads it and extracts clinical and administrative data at up to 99.5% accuracy in under two minutes. We match the patient, flag what is missing, and route a chart-ready referral into your EHR work queue.
Does AI referral management replace our intake staff?
No. It replaces the reading and keying, not the people. Staff review flagged exceptions, confirm unmatched patients, and handle payer edge cases, and humans stay on the exceptions by design.
Does it integrate with our EHR?
Yes. We integrate with Epic, athenahealth, eClinicalWorks, NextGen, AdvancedMD, DrChrono, Elation, Charm Health, ModMed, and Office Practicum. Your team keeps working in the EHR it already uses, and we write structured referrals into that work queue.
Does it handle prior authorization?
Prior authorization is a supporting workflow that feeds the same referral pipeline, not the headline. Our Prior Auth AI Agent captures structured data to keep authorizations moving. We label any prior auth performance figures as pilot-stage research targets, not production results.
How do we measure whether referral management is working?
Track the metrics you already report: referral leakage, time from referral received to appointment booked, staff hours reclaimed, and cost per document. Across our deployments, MagicFlow has driven a 40% reduction in referral leakage and 3x faster time to a specialist appointment.
Is your referral management HIPAA compliant?
Yes, we are HIPAA compliant, which means we protect patient health information under US federal privacy and security law. We are also SOC 2 Type II audited. We sign a Business Associate Agreement, the contract that governs how we handle your patient data, on paid tiers.
See the full referral workflow
Referrals, clinical documents, and fax run on one pipeline, and these resources show how the pieces connect.
- Automate inbound clinical faxes with our Fax AI Agent.
- Explore intelligent document processing inside MagicDocs.
- Read how AI fax services work for front-office teams.
- Follow the healthcare AI fax guide for specialty groups.
- See neurosurgery referral management in practice.
- See orthopedic referral to appointment end to end.
Ready to see it on your own referrals? Book time with our team.

Get Started
AI-Enhanced Healthcare with Insight Health


