TriFetch Raises $1.9M Pre-Seed to Bring AI Automation to Specialty Clinic Intake, Referrals, and Prior Authorizations

The front desk never makes the headlines, but it sets the tempo. Phones stack up, referrals stall out, prior auths drift into limbo, and somewhere in that noise, revenue leaks while clinicians wait to actually practice medicine. It’s not chaos, it’s routine, which might be the bigger problem.

TriFetch, out of San Francisco, just pulled in $1.9M in pre-seed funding led by Nexus Venture Partners, with a sharp bench of operators from Google, Hippocratic AI, Mercor, and MIT in the mix. Capital meets constraint, and suddenly the most ignored layer in healthcare starts getting VIP treatment. Administrative work isn’t sexy, but it’s expensive, and now it’s funded like it matters.

Varuni Sarwal, Ph.D., and Rosemary He, Ph.D., didn’t stumble into this. They met in the trenches of UCLA’s computer science PhD program, where AI isn’t a buzzword, it’s a tool you either sharpen or get cut by. One built models around EHR data, the other worked on medical imaging. Together, they looked at clinics and saw something obvious hiding in plain sight. The system isn’t just strained on the clinical side, it’s suffocating under its own paperwork.

But ideas don’t ship themselves. That next layer, the one that turns research into something a clinic can actually rely on at 4:47 PM on a Friday, runs through the people building it. Dhruv Miyani shaping the core platform for reliability at scale. Dhairya Shah engineering the backbone that keeps workflows moving without friction. Yashvardan Singh wiring the infrastructure so integrations don’t become another headache. Small team, heavy lift, no room for theory when the phone keeps ringing.

TriFetch steps in as the connective tissue. A multilingual voice agent that doesn’t just pick up calls, it handles them. Scheduling, follow-ups, the whole loop. A referral engine that routes, verifies, and books without turning into a game of telephone. Prior authorizations that don’t sit in limbo while patients wait and providers chase ghosts. And it plugs into systems like NextGen, eClinicalWorks, and Athena without asking clinics to rip out what already works.

The numbers tell their own story if you listen closely. Around 16 hours of staff time freed up per day for a mid-sized specialty clinic. Up to $1.4M a year in recovered costs and captured revenue. That’s not “efficiency.” That’s giving a clinic its pulse back.

What stands out isn’t just the tech, it’s the restraint. Humans stay in the loop where it matters. No chest-thumping about replacing staff, just a quiet understanding that the best systems don’t fight the workflow, they slide into it and make everything feel a little less chaotic.

Nexus Venture Partners didn’t just fund a product, they backed a pressure point. The kind that, once relieved, changes how the whole body moves. And TriFetch is starting with specialty clinics, but you can already see the outlines stretching toward multi-specialty groups, primary care, and hospital networks.

There’s a lesson buried in here for anyone building right now. The loud ideas get attention, but the ones that remove friction get paid. TriFetch didn’t invent healthcare. They just decided the back office shouldn’t feel like a black hole. And if you’ve ever been stuck on hold listening to elevator music while someone hunts for your paperwork, you already know why this matters.