The inefficiencies of the U.S. healthcare referral process became personal for Tennr cofounder Diego Baugh while he was in college. After being hospitalized for stomach issues, Baugh was referred to a specialist but weeks passed with no communication. He never got the appointment. A second hospitalization led to another referral and another six-week delay. It wasn’t until he was hospitalized a third time that he realized just how widespread and damaging the problem truly was.
That experience would later shape the foundation of Tennr, a health tech startup launched in 2021 by Baugh, Trey Holterman, and Tyler Johnson. The New York-based company was built to solve a deceptively simple but deeply systemic issue: referrals from primary care and urgent care to specialists are riddled with delays, denials, and data drop-offs.
Tennr’s mission to fix that has gained serious investor backing. Less than a year after closing a $37 million Series B, the company has now raised a $101 million Series C round, led by IVP, with participation from Google Ventures, Iconiq, and existing investors Andreessen Horowitz and Lightspeed. The new funding values Tennr at $605 million, making it one of the fastest-growing health tech startups in an increasingly AI-driven sector.
CEO Trey Holterman, who previously worked as a software engineer at Health IQ and Strava, credits the idea not just to Baugh’s referral nightmare, but also to his own mother, a family medicine physician who had long voiced frustration over the administrative chaos of the referral process. “I probably should’ve just listened to my mom sooner, which I think is a tale everybody knows,” Holterman said.
Fixing a System Few Have Tackled
Each year, roughly one-third of Americans are referred to specialists, but more than half don’t complete the next step. Referrals come in through faxes, emails, PDFs, and scanned documents where none of them are standardized. The burden of deciphering those materials falls on already-overwhelmed clinics and care coordinators, resulting in long delays for patients and often no follow-up at all.
Tennr set out to build an intelligent platform that could parse the web of paperwork, verify insurance eligibility, interpret doctors’ notes, and help shepherd patients from primary care to specialty visits.
Even before the release of ChatGPT in 2022, Tennr had begun building its own specialized machine learning model. Rather than relying on general-purpose AI tools, the founders decided to train their models on tens of millions of medical documents, with a focus on patient referrals. That strategy has remained core to the company’s differentiation.
Tennr’s proprietary model, called RaeLM, is trained on over 100 million medical documents, 2.3 billion distinct data fields, and 8,000 sets of payer rules and eligibility criteria. It’s purpose-built for the healthcare administration where generalized AI often fails to deliver reliable results. In the real world, eligibility checks can be wrong nearly 8% of the time, according to Tennr. RaeLM reduces those error rates by extracting key fields and automatically confirming benefit data using custom logic.
“We’re betting on open source and a proprietary dataset that we’ve accumulated, and that continues to totally smash benchmarks,” Holterman said. He emphasized that Tennr complies with HIPAA regulations through de-identification of medical records during model training.
Tennr also developed T3, or Transcript Translation Technology—a system trained specifically to handle the often chaotic and repetitive world of medical call transcripts. Using probabilistic models and domain-specific heuristics, T3 can process recorded phone calls, identify relevant data like member IDs, and convert those interactions into structured, actionable tasks.
Holterman is quick to stress that Tennr’s goal is not to automate away human jobs, but to ensure that patients actually get the care they’re referred to. “It’s really about making sure that the patient actually gets the service and understands what it’s going to cost, so that they show up,” he said.
From Weeks to Hours
Today, Tennr’s platform is used by a growing network of healthcare providers ranging from individual physicians to clinics and hospitals—that handle a high volume of referrals. The system is capable of ingesting documents in any format and standardizing them to trigger the next steps in the care journey. That includes eligibility verification, insurance lookup, appointment coordination, and provider-patient communication.
The impact has been measurable. Tennr has already facilitated tens of millions of patient referrals, many of which now result in specialist appointments within hours, rather than the weeks-long delays typical of the current system.
“We’re really excited about how many Americans we’re now helping behind the scenes,” Holterman said.
Early users of Tennr’s platform are already reporting major improvements in efficiency. “We’re now processing new patients in a fraction of the time it used to take, which has been a game-changer,” said Darius Reid, head of operations at Total Medical Supply, which works with patients in Texas. He called the platform “transformative to our workflow.”
A Strategic Wedge in Healthcare AI
IVP’s Zeya Yang, who led the Series C round, first backed Tennr during its Series A while at Andreessen Horowitz. He sees Tennr’s focus on referrals as a compelling entry point. “This can be a very big company if they figure that sort of stuff out,” Yang said.
Yang highlighted that Tennr has built a strategic wedge into the complex healthcare ecosystem that allows the company to serve not just specialists, but also primary care providers and, increasingly, patients. The company is now expanding its offering to include network visibility tools that help both physicians and patients track the status of referrals and payments in real-time.
Tennr’s financial growth reflects that momentum. While the company has not disclosed exact revenue figures, Holterman confirmed that Tennr is operating in the eight figures, and revenue has tripled since the Series B raise in October 2023.
The company plans to double its headcount over the next six to eight months and is continuing to invest in new capabilities including better automation for processing complex healthcare codes.
Despite the company’s extensive use of AI, Holterman remains cautious about branding Tennr as another AI-first startup. “I want to talk about problems and I want to talk about solutions,” he reiterated. “I don’t want to talk about just the technology.”
As for what’s next, Holterman said the company is staying focused on execution. “We see a real opportunity to step on the gas and be as aggressive as possible,” he said.
