Healthcare AI is having a moment, and Tennr might be the clearest example of how fast the right company can scale in this space.
The New York-based startup raised a $101 million Series C in June 2025, led by IVP with participation from Google Ventures, ICONIQ, Andreessen Horowitz, Lightspeed, Foundation Capital, and Frank Slootman (former CEO of Snowflake). The round values Tennr at $605 million and brings total funding to $162 million.
The company closed a $37 million Series B less than a year earlier. And according to CEO Trey Holterman, they hadn't even touched that Series B cash yet when the C came in. When your investors are preempting your next raise before you've spent the last one, it tells you something about the traction.
Here's a scenario that plays out millions of times a day in American healthcare. A primary care doctor refers a patient to a specialist. That referral gets faxed (yes, faxed) to the specialist's office along with a stack of clinical notes, insurance forms, and prior authorization documents. Someone at the receiving end has to manually read through all of it, figure out if the patient's insurance will cover the visit, check whether the right documentation is in order, and schedule the appointment.
It's slow. It's error-prone. Patients fall through the cracks constantly. Referrals get lost, prior auths get denied, and people who need care don't get it on time. Holterman learned about this problem from his mother, who worked in family medicine and showed him firsthand how chaotic the handoff between providers actually is.
Tennr automates that entire workflow. Their AI reads the incoming documents (including handwritten doctor's notes, which is a genuinely hard computer vision problem), checks them against payer criteria, flags potential denials before they happen, and routes everything to the right place. The goal is to get patients from referral to first visit faster, with fewer denials and less manual work for office staff.
Tennr built RaeLM, a proprietary vision-language model trained on over 100 million anonymized healthcare documents and 2.3 billion distinct data fields. It's purpose-built for healthcare intake, which matters because general-purpose LLMs from OpenAI or Anthropic aren't optimized for this kind of work. Parsing a faxed referral with a doctor's handwriting, checkboxes, scanned forms, and abbreviations specific to a particular specialty requires a model that's been trained specifically on that data.
Holterman has been vocal about this. He's said that betting on a proprietary dataset continues to "totally smash benchmarks" compared to generic models, even as foundation models get better. The reasoning is straightforward: the volume of hyper-specific healthcare documentation data that Tennr has accumulated would never make economic sense for a broad AI lab to chase. It's too niche and too messy. That's exactly what makes it defensible.
The company also built the system to comply with HIPAA through de-identification, which is table stakes for healthcare AI but worth noting given how much sensitive data flows through the platform.
Tennr now processes over 10 million documents per month. The company works with hundreds of healthcare organizations hosting thousands of providers. Revenue is in the eight figures and tripled from the Series B level in under a year. The team has grown to 256 employees as of January 2026.
For context, Tennr came out of Y Combinator's Winter 2023 batch. Going from a YC startup to a $605 million valuation in about two years is a trajectory that puts them in rare company, especially in healthcare, where sales cycles are long, compliance is heavy, and trust takes time to build.
IVP partner Zeya Yang, who led the Series C, pointed to something specific about Tennr's positioning. The referral management problem looks niche from the outside, but it's actually a wedge into a much larger market. Every specialist provider deals with this. Orthopedics, cardiology, dermatology, oncology. The documentation and intake challenges are universal, even if the specific criteria vary by specialty and payer. Tennr's playbook is to nail referral automation, then expand outward into adjacent parts of the healthcare operations stack.
Alongside the funding announcement, Tennr launched Tennr Network, which connects referring providers, receiving providers, and patients on a shared platform for real-time referral visibility.
This is a meaningful product expansion. Previously, Tennr's AI handled the document processing on the receiving end. The network adds a layer of connectivity that gives everyone involved (the doctor who made the referral, the specialist receiving it, and the patient) visibility into where things stand. Did the referral go through? Is the prior auth approved? What's the financial responsibility going to look like?
Holterman has talked about how even with a perfectly optimized operation, patients still fall through the cracks because they lack visibility into their own referral status. They don't know what's happening, they don't understand what it's going to cost, and they end up no-showing or delaying care. Tennr Network is designed to close that gap.
Healthcare AI funding was enormous in 2025. Abridge raised $550 million across two rounds. Ambience Healthcare pulled in $243 million. OpenEvidence raised $210 million. The money flowing into the sector reflects a real belief that AI can fundamentally change how healthcare operations work.
But healthcare is also where AI companies go to learn hard lessons about implementation. Sales cycles are long. Buyers are risk-averse. Regulatory requirements are strict. Integration with legacy systems (many of which still run on fax machines and decades-old EHRs) is genuinely painful.
Tennr's advantage is that they started with a specific, well-defined problem that every specialist practice deals with. They didn't try to boil the ocean with a general "AI for healthcare" pitch. They went after referral automation, built a model purpose-trained for it, proved it works at scale, and are now expanding from that foundation.
That focused approach is also what attracts a specific kind of investor confidence. Frank Slootman backing the round is notable. The former Snowflake CEO is known for investing in companies with clear operational leverage and a path to becoming essential infrastructure. That's the trajectory Tennr is aiming for: becoming the default system that sits between referring providers and specialists, processing the messy paperwork that makes healthcare run.
Tennr's Series C cash will go toward expanding the engineering team, scaling go-to-market efforts, and building out the Tennr Network. The company is also continuing to deepen its model's capabilities across more specialties and payer criteria sets.
At 256 employees and eight figures in revenue that's tripling year over year, Tennr is past the "promising startup" phase and into genuine scale-up territory. The question now is how far the referral wedge can take them, and whether the Tennr Network becomes a connective layer that healthcare systems rely on the way they rely on their EHR.
If the last year is any indicator, they're going to move fast. They went from Series B to a $605 million valuation Series C in under twelve months. In healthcare AI, where everything is supposed to take forever, that pace speaks for itself.