The AI healthcare admin worker that logs into insurance portals, reconciles the money, and posts it - so the doctor doesn't have to.
Here is a fact about running a dental office that nobody puts on a brochure: a surprising share of the day is spent logging into insurance portals, downloading remittances, matching them line-by-line against what the practice thinks it is owed, typing the results into a practice-management system, and then confirming the money actually landed in the bank. It is repetitive. It is unforgiving. And a practice can spend somewhere between $100,000 and $200,000 a year to have humans do it.
Lassie's pitch is that this is a software problem wearing a staffing costume. The company builds an AI agent that does the work - not a dashboard that tells you about the work, not a copilot that drafts a first pass for a human to approve, but an agent that goes into the portal, pulls the reimbursement, reconciles it, posts it to the system-of-record, and verifies the funds. The company says it handles roughly 98% of posting autonomously and flags the remaining slice for a human, which is the correct and honest ratio to advertise, because the 2% is where the judgment lives.
That distinction - do the work, don't add to it - is the whole personality of the company. Plenty of AI tools quietly create a new job, the job of checking the AI. Lassie is built to remove a job instead. When the founders describe the product, the verbs are all things a biller would recognize: enroll, reconcile, post, appeal, follow up, report. When the marketing gets sentimental, it lands on the line "You're a doctor. Not a machine." - which is a tidy way of saying the machine is supposed to be the software, not the person at the front desk.
The company started with dental practices, which is a smart place to start a business you eventually want to be much larger. Dental billing is standardized enough to automate, painful enough to pay for, and referral-friendly enough that one office manager will happily tell another. Lassie grew mostly on exactly that: referrals and inbound demand, the kind of growth that is slow to start and hard to fake.
"Businesses will soon run themselves, and small businesses will be among the first." - Andreessen Horowitz, on why it led the round.
"A breath of fresh air for your office."
- Lassie's promise to the front desk
Enrolls electronic payments, reconciles EFTs against records, and posts them to the practice-management system - the core loop, run at ~98% autonomy.
Handles insurance enrollments, claim appeals, follow-up tasks and reporting, so recurring admin doesn't pile up on a human's desk.
Answers practice questions, gives status updates, and surfaces the small share of cases that genuinely need a person to decide.
Before there was much of a product, the founders reportedly visited on the order of 100 dental offices - sitting where the busywork happens and watching how billing actually gets done. That field-research habit is the tell. It is why the product reconciles an EFT the way an experienced biller would, and it is the sort of thing a term sheet can't manufacture after the fact.
An early product manager at Robinhood and Coinbase, Pelle brings consumer-fintech instincts - trust, money movement, and the unglamorous plumbing behind both - to a healthcare back office.
The first product hire at Superhuman, with time at Uber, Renken carries a craft-and-detail obsession into a category not exactly known for polish.
The angel list reads like a fintech and product-craft hall of fame - the founders of Plaid, Wise, Superhuman, and Reforge - which fits a company whose whole thesis is that money movement plus product obsession can automate a workflow everyone else finds too dull to touch. Reported metrics at the raise: $10M+ annualized revenue, most of it from referrals.
Founders embed in dental practices - visiting roughly 100 offices - to learn billing and reconciliation firsthand.
Lassie reaches 700+ practices in 49 states, delivers 250,000+ labor hours a year, and passes $10M in annualized revenue.
Andreessen Horowitz leads a $35M round to push the autonomous admin worker beyond dental.
There is a version of the AI story that is all robots and factory floors, and there is a quieter version where the real change is an agent closing the books at a dental office in Ohio while everyone sleeps. Lassie is a bet on the quiet version. The company - and its lead investor - frame the mission as small businesses that increasingly run themselves: not just billing, but eventually answering phones, scheduling, hiring, collecting payments, and handling the exceptions in between.
Dental is the wedge, not the destination. The appeal of starting there is that it forces the hard, boring discipline of being right about money, in a regulated (HIPAA) environment, at five-figure annual contract values that a practice will actually pay because the alternative costs more. If Lassie can be trusted to touch a practice's reimbursements without creating a cleanup job, the same trust is portable to the next workflow, and the one after that.
The competition is mostly the status quo: in-house billing staff, legacy revenue-cycle outsourcers, and a growing field of AI-agent startups eyeing healthcare administration. Lassie's edge, for now, is that it measures itself in hours returned rather than insights generated - 250,000 of them a year - which is a refreshingly literal way to keep score. Whether that scales from 700 practices to the broader small-business back office is the open question, and it is the one worth watching.
For a company named after a loyal dog, the promise is fittingly modest and specific: fetch the paperwork, bring back the money, don't make a mess. In a category prone to overpromising, doing one dull thing extremely well is a reasonable place to build from.
It builds AI agents that automate back-office admin for independent medical and dental practices - logging into insurance portals, pulling reimbursements, reconciling and posting payments, handling appeals and enrollments, and flagging exceptions for staff.
Steijn Pelle (CEO), an early product manager at Robinhood and Coinbase, and Frédéric Renken, the first product hire at Superhuman.
A $35M Series A led by Andreessen Horowitz in June 2026, bringing total funding to roughly $47M.
700+ practices across 49 states, delivering over 250,000 labor hours to customers each year.
Yes. Lassie is HIPAA-compliant and designed to handle sensitive practice and payment data, flagging exceptions for human review.
Figures (funding, practice count, revenue, labor hours) are drawn from public reporting and company statements as of mid-2026 and are approximate. Founding year not publicly confirmed.