Assort Health’s AI Aims to Plug Healthcare’s Leaky Referral Funnel

SAN FRANCISCO, CA – August 12, 2026 – For countless patients, it’s a familiar and frustrating journey into a communications black hole. A primary care doctor issues a referral to a specialist, a slip of paper or a digital message is sent, and then… silence. Days, sometimes weeks, pass without a call. For the specialty practice on the receiving end, that referral is one of dozens, perhaps hundreds, that arrive daily via a chaotic mix of faxes, EHR messages, and scanned PDFs, burying staff in a mountain of administrative work. This costly inefficiency, known as referral leakage, represents a significant loss of revenue for providers and a dangerous delay in care for patients.

Now, San Francisco-based Assort Health, fresh off a $120 million Series C funding round that values the company at $1.2 billion, is rolling out a solution that aims to automate this broken process from end to end. The company today announced the general availability of Referrals, an AI agent designed not just to read and sort documents, but to actively manage each referral to completion. The business implications are profound: Assort claims its platform converts 79% of referrals into scheduled appointments, a figure that would represent a seismic shift in patient acquisition and revenue protection for beleaguered health systems.

From Fax Machines to Autonomous Agents

The traditional referral workflow is a relic of a pre-digital era, a manual process that even modern Electronic Health Records (EHRs) have failed to fully solve. Staff must manually extract information, verify insurance eligibility, create or update patient charts, and then begin a multi-step game of phone tag to schedule an appointment.

"Before Assort, our referral team was overwhelmed by the volume of incoming referrals," said Jensine Joseph, Call Center Manager at the Center for Sports Medicine & Orthopaedics (CSMO). "With roughly 75 referrals coming in each day, it could take three or four business days before a patient received their first outreach. Follow-up attempts sometimes didn't happen for a week or more."

Assort’s Referrals agent replaces this manual slog with an autonomous workflow. The AI ingests referrals from any source, extracts the relevant data, and applies the practice's specific business rules—from insurance acceptance to provider-specific criteria. It then proactively contacts the patient via their preferred method—voice, text, or email—to confirm details, resolve any missing information, and book an appointment directly into the EHR.

The results at CSMO have been transformative. "Today, Assort automatically processes referrals, verifies key information, and proactively reaches out to patients within hours of receiving the referral," Joseph stated. "That speed has made a tremendous difference. Patients hear from us before they start looking for other options, and we're now automating about 90% of referrals end-to-end and seeing a 77% conversion rate."

The Business Case for Intelligent Automation

The financial stakes are enormous. The administrative burden in U.S. healthcare is estimated to be a staggering $1.1 trillion problem annually. Referral leakage is a major contributor, with stalled or abandoned referrals representing lost appointments and downstream revenue. By automating the process, Assort Health is making a direct appeal to the bottom line of specialty practices and health systems whose growth depends on a steady stream of new patients.

"Healthcare organizations do not need another inbox or dashboard that asks their staff to manage the same work in a different place," said Jeffery Liu, Founder and Co-CEO of Assort Health. "Assort's Referrals agent does the work itself... That helps organizations convert more referral demand into scheduled care, protect a critical source of patient volume, and allow staff to focus on the cases that truly require their attention."

The company reports that its platform can drive up to $4.3 million in new revenue capture per 100 providers. This isn't just theoretical; clients like Michigan Orthopedic Surgeons have reported generating $2.3 million in additional revenue after implementing the system. This focus on tangible ROI is a key reason why investors, led by Menlo Ventures, have poured over $222 million into the company.

"A fax getting read isn't the same as a patient getting care," added Jon Wang, Founder and Co-CEO. "A stalled referral rarely announces itself. The patient just doesn't show, and the practice never learns why... We don't count a referral done until the patient is actually sitting with the specialist." This philosophy of seeing the process through to completion—from document to appointment—is a core part of Assort’s value proposition.

The Rise of the Specialized AI Brain

What sets Assort’s platform apart, according to the company, is its proprietary AI model, Synapse. While general-purpose AI models are powerful, they often lack the nuanced understanding required for healthcare's complex and highly regulated environment. Synapse, by contrast, was built from the ground up on a massive dataset of over 225 million specialty-specific patient interactions, care protocols, and decision pathways.

This specialized training allows the Referrals agent to understand the intricate rules of specialties like orthopedics, ENT, and gastroenterology. It can correctly match a patient with a knee complaint to a surgeon who specializes in that specific joint, factoring in insurance, location, and the provider's schedule. In fact, Assort claims that in 50% of cases, its AI identifies a better-fit provider than the one originally suggested, optimizing care and ensuring physicians see patients they are best equipped to treat.

Furthermore, the platform features what Assort calls "Patient Journey Memory." This enables a continuous, context-aware conversation. If a referral outreach goes unanswered and the patient calls the office two days later, the AI answering the phone knows about the pending referral and can pick up the conversation seamlessly. This eliminates the common patient frustration of having to start over with each interaction and builds a more cohesive, less fragmented journey to care.

Reshaping the Healthcare Frontline

The introduction of such powerful automation inevitably raises questions about the future of the healthcare workforce. Rather than aiming for outright replacement, however, the technology appears to be reshaping the roles of patient access staff and referral coordinators. By automating the 90% of routine, repetitive tasks, the AI frees human staff to function as expert problem-solvers for the remaining 10%—the clinically complex cases, the emotionally distressed patients, and the exceptions that require human judgment.

Cases that require clinical review or manual intervention are automatically routed to the appropriate staff member with all relevant information attached. This transforms the role from a data-entry clerk into a high-value case manager. According to industry analysts, this shift not only improves operational efficiency but can also increase job satisfaction and reduce burnout among administrative staff, a critical benefit amid widespread healthcare labor shortages.

By providing clear analytics on referral sources, conversion rates, and payer mix, the platform also elevates referral management from a back-office function to a strategic tool for business development. Practices can gain unprecedented insight into which referring relationships are most valuable and where bottlenecks are occurring, allowing them to protect the very channels that sustain their growth.