Forus Just Raised $160M at a $1B Valuation to Become the Operational Routing Layer for Specialty Rx & the Real Story Is Why Investors Now Believe Healthcare Orchestration Beats SaaS
Video Preview
🎧 Part I Podcast free on Spotify.
🎧 Part II Podcast episode for paid subscribers only. Also available on Spotify.
To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).
Abstract
Healthcare is structurally a transaction routing economy, not a care delivery industry, and the most valuable real estate sits at the seams between stakeholders
Forus raised over $160M ($123M of it as Series B per AlleyWatch) led by Thrive Capital and General Catalyst at a $1B valuation on roughly $50M of expected annualized revenue this year
The thesis being underwritten is operational and infrastructural, not a workflow software bet
Specialty drug economics changed the math by making every coordination failure expensive enough to justify rebuilding the plumbing
AI collapses the labor cost of bridging fragmentation, which is the historical reason orchestration was never commercially viable at scale
Traditional digital health SaaS keeps plateauing because seat licenses cannot capture transaction value and because healthcare workflows fragment continuously rather than standardize
Forus monetizes biopharma manufacturers downstream while giving the product away free to providers and patients, which is the same business model as Visa and Google
The incumbent stack of PBMs, hubs, specialty pharmacies, and EHR vendors is going to fight this category hard
The orchestration layer compounds on multi sided network effects rather than feature parity
The next decade of healthcare infrastructure outcomes will probably look more like Stripe than like Veeve
Table of Contents
Why a Series B Round Is Actually a Market Signal
The Coordination Economy Hiding Inside Healthcare
How Specialty Drugs Quietly Rewired the Admin Layer
Why Most Digital Health SaaS Plateaued
What Forus Actually Built and Why It Works
The Math on Free Software That Still Generates Real Revenue
Why AI Is the Cheat Code for Fragmentation
The Looming Control War Between PBMs Hubs and Orchestrators
What Investors Are Really Underwriting
Final Thoughts on the Routing Layer Thesis
Why a Series B Round Is Actually a Market Signal
On May twelfth, a New York company most healthcare operators had not heard of two weeks earlier announced it had pulled together over $160 million in equity financing at a $1 billion valuation. The company is Forus, previously branded as Tandem, founded in 2023 by a thirty one year old named Sahir Jaggi who used to be a partner at Rough Draft Ventures and earned a biomedical research degree at Columbia. The round was led by Thrive Capital and General Catalyst, with Accel, Bain Capital Ventures, Redpoint, BoxGroup, and Pear VC alongside. Roughly $123 million of that came in as the new Series B per AlleyWatch’s reporting, and the company is expecting to clear $50 million in annualized revenue this year.
On its face this looks like another late cycle digital health raise, of which there have been plenty in the last eighteen months, most of them now quietly being forgotten by the same LPs who got pitched on them. What makes this one different is not the headline number, not the cap table, and not the now mandatory “AI powered” descriptor in the press release. It is the underlying thesis, which is operational rather than technological. The investors did not write a check because they believe in another point solution that automates prior authorization or another patient access platform built on a workflow tool. They wrote a check because they believe the bottleneck in American healthcare is no longer storage, integration, or analytics. The bottleneck is transaction coordination, and whoever owns the orchestration layer in the next decade is going to own something closer to Visa than to Veeva.
That last part deserves to be stated plainly. The market has spent fifteen years assuming healthcare looks like enterprise software, where the winners are the companies with the cleanest interfaces and the deepest workflow penetration in a single user persona. What is actually happening, and what this round reflects, is that the most valuable real estate in healthcare is the routing layer between stakeholders, not the application layer sitting on top of any one of them. That is a very different game with a very different power law, and the people writing the checks at this stage are starting to act like they finally figured that out.
The Coordination Economy Hiding Inside Healthcare
Most outsiders think healthcare is a care delivery industry. It is not. Healthcare is, mechanically, a transaction routing economy that occasionally performs medicine on the side. The reason this distinction never quite lands is that the routing is invisible to the patient and largely invisible to the physician, but it is doing essentially all of the real work between the moment a clinical decision gets made and the moment a therapy reaches the body.
Every referral is a transaction. Every prior authorization is a transaction. Every benefits verification is a transaction. Every formulary check, every specialty pharmacy assignment, every copay assistance enrollment, every reimbursement determination, every credentialing verification, every step therapy override, every appeal, every shipment, every refill authorization, every adherence outreach, every patient assistance program eligibility recalculation. These are routine operational events that happen by the tens of millions every week, and the entire economy of intermediaries in healthcare was built to handle the friction at the seams between them.
PBMs exist because the system is fragmented. Clearinghouses exist because the system is fragmented. Revenue cycle vendors exist because the system is fragmented. Specialty pharmacies exist because the system is fragmented. Provider data companies exist because the system is fragmented. Hub services companies exist because the system is fragmented. Prior authorization vendors exist because the system is fragmented. If you collapsed the fragmentation, you would erase trillions of dollars of intermediary revenue in the process, which is also why no incumbent has any economic incentive to collapse it.
The companies that have historically gotten genuinely large in healthcare technology did not do it by building the prettiest dashboards. They did it by inserting themselves into transaction flow and monetizing volume. CoverMyMeds got bought by McKesson for roughly $1.4 billion in 2017 because it had become the de facto rails for electronic prior authorization, not because it had a beautiful front end. Surescripts is a network, not a product. Change Healthcare was, fundamentally, a clearinghouse before UnitedHealth swallowed it for around $13 billion. The pattern is the same in every category that has ever produced an outlier outcome in this industry. Software seats are a tax. Transaction control is a toll. The difference compounds.
How Specialty Drugs Quietly Rewired the Admin Layer
The reason this is finally hitting an inflection point in 2026 instead of in 2010 has almost everything to do with what happened to the drug pipeline over the past decade. When most therapies cost a few hundred dollars a month, administrative friction was a nuisance but it was survivable. The system tolerated phone calls and faxes and three day turnarounds because the dollars at stake on any individual prescription were not large enough to justify rebuilding the plumbing. That tolerance ended the moment specialty became the dominant share of pharmacy spend.
A modern specialty therapy can cost anywhere from sixty thousand dollars a year to three million dollars one time across the course of treatment. Cell and gene therapies, biologics for inflammatory and rare disease, oncology infusions, GLP ones in their high cost indications, and the wave of pipeline assets behind them all sit in this band. When the average prescription on a patient population is that expensive, every operational failure carries an enormous shadow cost. A prior authorization denied for a missing piece of clinical documentation can delay therapy by weeks and cost the manufacturer a fully written script. A benefits verification routed to the wrong specialty pharmacy can blow up patient onboarding entirely. A copay enrollment that fails because the hub did not communicate cleanly with the foundation can collapse adherence in the first ninety days, which is exactly the window where specialty persistence is most fragile.
What changed is that the transaction layer itself became economically valuable in a way it simply was not when most prescriptions were generic statins and ace inhibitors. Manufacturers will now pay almost any reasonable price for a service that reliably shortens time to therapy by a week, because the net present value of an additional patient on a sixty thousand dollar a year biologic dwarfs the per patient cost of the service. Providers will adopt almost any tool that reduces the staff hours burned per prescription, because they are losing money on every minute their nurses and pharmacists spend on hold with payers. Payers quietly tolerate anything that drives better clinical documentation upstream because their own utilization management economics depend on it. The entire stakeholder set is finally aligned in a way it was not for the previous twenty years, and that alignment is what makes a true orchestration layer commercially possible for the first time.
Why Most Digital Health SaaS Plateaued
The graveyard of digital health is full of beautifully designed products that hit a wall somewhere between thirty and seventy million in annual recurring revenue and then never broke through. The companies kept telling their boards the next big customer was around the corner, the boards kept telling their LPs the same story, and the customers kept slow rolling the procurement until eventually the cap table got pushed underwater by a flat round and the founders went to go do something else.
The reason this happened so consistently is structural. Traditional enterprise SaaS scales because workflows standardize over time, which lets one product address more and more buyers with the same code base. Healthcare workflows do the opposite. They fragment continuously because reimbursement methodologies change, payer rules change, formulary positions change, clinical pathways change, network configurations change, benefit designs change, regulatory requirements change, and specialty drug economics evolve faster than software roadmaps can possibly track. Every standard you ship is obsolete within twelve months because the ground has moved underneath it.
The companies that survived this dynamic almost always did it by abandoning the seat license model. They moved to transaction based pricing, hybrid models, capitated arrangements, percentage of spend, or revenue share with downstream stakeholders. The companies that stayed pure SaaS struggled because the buyer kept rebudgeting them as discretionary spend even when the product was beloved by end users. Health systems are not software companies. They do not allocate budget like software companies. They allocate budget the way operations heavy enterprises do, which means anything that does not show up as either revenue or cost takeout eventually loses funding when finance gets squeezed.
The other thing the SaaS playbook never solved is the network problem. A patient access platform is only useful if the manufacturer and the provider and the pharmacy and the hub are all reachable through the same connection. If the platform only sits with the provider, the manufacturer still has to maintain its own hub. If it only sits with the manufacturer, the provider still has to do everything manually. The economics work only if you control the connections on multiple sides at once, which is to say if you are running a network rather than a product. Networks are much harder to build, take much longer to compound, and require a category of patience that public market software investors have basically run out of. That patience now sits almost entirely with crossover and growth equity firms who are willing to underwrite ten year compounding stories at infrastructure economics, which is exactly the firm profile leading this Forus round.
What Forus Actually Built and Why It Works
Forus is, structurally, an orchestration network embedded directly inside the physician workflow at the point of prescribing. The moment a clinician writes a prescription, the system picks it up and handles essentially everything that has to happen between that click and the patient actually starting therapy. It checks medication history, identifies pharmacy restrictions, manages prior authorization documentation, routes the prescription to the right fulfillment pathway, and gives both the prescriber and the patient real time visibility into where the process is stuck. The reported result, per the company’s own materials and external coverage, is that approvals that historically took months are now closing in two to three days, and median time to therapy initiation has compressed from over a week to roughly one day.
The product is given to providers and patients for free. The revenue model is downstream, captured through relationships with biopharma manufacturers, who pay for the orchestration because they get the thing they want most in the world, which is patients on therapy faster and more reliably. Five of the top ten global biopharma companies are reportedly already commercially engaged, and the company expects to clear $50 million in annualized revenue this year on what is functionally a zero priced product to the demand side of the market. That structure is the actual unlock. By giving the product away to the side of the equation that has the budget squeeze, Forus solves the historical adoption problem that crushed every prior authorization SaaS company. By monetizing the side of the equation that experiences the operational ROI most directly, it has a willingness to pay curve that scales with the value of the prescription rather than with the number of users.
The adoption signal is the part that should make competitors uncomfortable. Provider penetration grew ten times year over year for the past two years and the company now claims coverage of roughly eighty percent of US residential zip codes, with thousands of medical practices and health systems across all fifty states actively using the platform. That growth has reportedly been driven almost entirely by word of mouth, which is the single hardest go to market metric to fake in healthcare. A health system office manager telling her counterpart at another system that something actually works is, empirically, the only marketing channel in this industry that consistently converts. The fact that this growth happened under the previous Tandem brand without a particularly aggressive PR posture is itself a tell about how much operational ROI the product is delivering on the ground.
The Math on Free Software That Still Generates Real Revenue
The economic structure of the business is worth sitting with for a second, because it explains a lot about why these particular investors moved at this particular valuation. A product priced at zero to the user but monetized through transaction adjacent revenue from a different stakeholder is the same basic business model that built Google, Visa, and most of modern advertising infrastructure. The thing that makes it work in any of these categories is that the side of the market paying the bills cares dramatically more about volume and quality of transactions than about the unit price of the underlying interaction. A pharmaceutical brand spending billions a year on field force, hub services, and patient support programs is not going to flinch at paying a meaningful fee per script started if the alternative is losing patients to abandonment.
To put numbers on it, consider what a typical specialty hub services contract looks like today. Manufacturers commonly spend somewhere in the range of one thousand to three thousand dollars per patient enrolled on bespoke hub operations, often more for therapies with complex enrollment requirements. They will spend that money on call center staffing, benefits investigation, prior authorization support, copay administration, and adherence follow up, and they will spend it even if the patient never actually starts therapy, because the unit economics of the brand still work out when persistence is high enough. If an orchestration network can deliver equivalent or better outcomes at meaningfully lower cost per enrolled patient while increasing speed to therapy and conversion, the manufacturer is buying a better outcome at the same or lower price. The orchestration company is capturing a fraction of what was already being spent, but doing it across many manufacturers at once with a software cost structure rather than a labor cost structure.
The reason this is so attractive at $50 million in annualized revenue with a billion dollar valuation is that the unit economics get dramatically better as scale builds. Every new manufacturer relationship increases the value of the network to existing providers because more therapies can be routed through the same pipes. Every new provider adoption increases the value of the network to existing manufacturers because more potential prescribers are reachable through the same connection. Every transaction generates routing intelligence that improves every subsequent transaction. The compounding is, in the technical sense, multi sided, and multi sided networks tend to either lose or win enormously. There is rarely a middle outcome.
Why AI Is the Cheat Code for Fragmentation
The reason this approach was not commercially viable a decade ago is that the operational layer in healthcare has always required a brutal amount of human labor to bridge the gaps between systems that cannot speak to each other. Benefits verification teams sit on hold with insurance companies. Prior authorization coordinators ferry clinical documentation between EHRs and payer portals. Hub case managers manage spreadsheets and faxes and PDF intake. Specialty pharmacy liaisons chase missing information across multiple systems. Patient access teams reconcile mismatched data across formularies and copay programs. The whole apparatus runs on a labor force whose existence is essentially a tax that the rest of the system pays to compensate for the fact that the underlying data and decision pathways are not connected.
What has changed is that large language models are unusually well suited to exactly this kind of work. They are good at reading unstructured clinical documentation. They are good at parsing poorly designed payer portals and faxed forms. They are good at extracting structured fields from ugly inputs. They are good at routing decisions where the rules are nominally written down but practically applied with a lot of edge cases. They are good at communicating in semi formal language with humans on the other side of a phone or an email. Almost every job that exists today in the healthcare administrative layer is, in some real sense, a job that a properly orchestrated AI system can do faster, more cheaply, and at higher consistency than a human can.
The thing to be careful about is that automating healthcare administration is not actually a model problem. The model is the easy part. The hard part is the orchestration around the model, the integrations into payer portals and EHRs and hub systems, the regulatory and compliance overlay, the human in the loop logic for edge cases, the change management with provider offices, and the actual operational rigor required to run a system that touches care delivery. Any reasonably capable foundation model can pass a prior authorization criteria check in isolation. Only a company with several years of integration depth and operational learnings can actually orchestrate the end to end workflow at scale. That moat is built operationally, not algorithmically, which is why investors with deep healthcare experience are placing bets on companies that have been quietly building in this space for years rather than on the freshly minted “we are an AI agent for prior auth” startups that have been pitching seed rounds for the past eighteen months.
The Looming Control War Between PBMs Hubs and Orchestrators
The category Forus and a handful of similar companies are building toward sits directly on top of territory that existing intermediaries currently consider theirs. PBMs control formulary and routing logic. Specialty pharmacies control fulfillment. Hub services vendors control patient enrollment and adherence. EHR vendors control the point of prescribing. Manufacturers control brand strategy and patient support program economics. Each of those stakeholders has historically built proprietary infrastructure to handle the slice of the workflow they touch, and each of them has an interest in not letting a neutral third party become the connective tissue across all of them.
This is going to get ugly before it gets stable. PBMs are not going to enjoy having an external orchestration layer that can see across formulary decisions and route around restrictive utilization management when clinical evidence justifies it. Specialty pharmacies are not going to enjoy losing some of their selection power to algorithmic routing based on patient geography, network status, and turnaround time. EHR vendors are going to want to build or buy the orchestration layer themselves rather than cede the in workflow real estate to a third party. Hub services vendors are going to feel directly threatened, because they are essentially the manual version of what Forus is trying to automate, and they know it.
The dynamic that often gets missed in conversations about this category is that the incumbents have most of the data but very little of the workflow leverage at the point of prescribing. The EHR has the workflow but not the cross stakeholder routing logic. The PBM has the formulary but not the prescriber relationship. The hub has the patient relationship but lives in a parallel universe from clinical care. The specialty pharmacy has the fulfillment but is downstream of every meaningful decision. An orchestration company that is embedded at the moment of prescribing, with relationships across manufacturers and pharmacies and payers, is structurally positioned to mediate among the rest of them, which is exactly the position that lets a Visa style intermediary capture a small percentage of an enormous transaction flow.
The category that loses most directly in this future is the manual hub services business. There is a real possibility that the field collapses meaningfully over the next five years as the orchestration layer absorbs the workflows that previously required outsourced staffing. The same is true of significant pieces of the patient access and prior authorization point solution market, where many of the companies that raised growth rounds in the 2021 boom are now stuck with products that do part of what an orchestration network does for free as a feature. Some of those companies will get acquired into the orchestration stack. Some will get rolled up by hubs trying to defend their turf. Most will quietly run out of runway over the next two to three years as the category consolidates around a small number of network winners.
What Investors Are Really Underwriting
The cap table tells you most of what you need to know about what is actually being underwritten here. Thrive and General Catalyst do not usually lead late stage rounds in companies whose ambition is to be a workflow vendor. Their checks scale with their conviction in network effects and infrastructure economics. Accel and Bain Capital Ventures have spent the past two decades watching network businesses outperform application businesses on every meaningful long term metric. Redpoint, BoxGroup, and Pear were earlier participants and have been on the inside watching the operational metrics develop quarter by quarter. None of these firms are paying a billion dollar valuation on $50 million in annualized revenue because they think Forus is a really good prior authorization product. They are paying it because they think Forus is becoming the routing layer for specialty prescriptions, and if that thesis plays out the multiple on the current revenue is going to look comically cheap in retrospect.
The comparison set is also worth dwelling on. Most of the digital health companies that went public in the last cycle traded at one to four times revenue inside two years of listing because the market correctly identified that they were workflow point solutions rather than infrastructure plays. Twenty times forward revenue at the Series B is a price that investors only pay when they believe the company is going to compound at infrastructure economics over a decade or longer. The right reference set is not other patient access companies. It is the early stage valuations of companies like Stripe and Plaid, which similarly looked overpriced on near term revenue when they raised growth rounds and which similarly turned out to be priced correctly on the basis of routing layer optionality.
The risk in this thesis is not that the operational opportunity is wrong. The opportunity is obviously real and the math on automating administrative friction in specialty therapy is uncontroversial. The risk is execution against an incumbent stack that is going to fight back, sometimes with regulation, sometimes with exclusivity arrangements, sometimes with parallel internal builds, and sometimes simply with the inertia of healthcare contracting. Whoever wins the orchestration category is going to need to be diplomatically excellent with PBMs and pharmacies and EHRs while simultaneously eating their lunch operationally, which is not a particularly easy combination of qualities to find in a single management team. The early signals from the Forus go to market suggest the team understands this, but it is still very early.
Final Thoughts on the Routing Layer Thesis
The broader thing to take away from this round, regardless of what one thinks of Forus specifically, is that the center of gravity in healthcare infrastructure investing has clearly moved. For the past decade, the most ambitious money was chasing data plays, analytics plays, interoperability plays, and clinical AI. Those categories are not going away, but the next decade of giant healthcare technology outcomes is going to be defined by who controls the operational layer that moves prescriptions and authorizations and patient flows between fragmented stakeholders. That control is more strategically valuable than any individual application or dataset, because it sits on top of every meaningful transaction in the system.
The competitive set is going to expand quickly. Lumini, Qventus, Confido Health, and a handful of others are already pursuing variations of this thesis from different angles, and the AI native cohort raising seed rounds today is going to produce a few additional credible entrants over the next twelve to eighteen months. Some of the incumbents will respond with acquisitions, some with internal builds, some with strategic partnerships, and some with the kind of contractual brick walls that PBMs have historically deployed to slow down disintermediation. The market will sort itself out faster than most participants are expecting, because the underlying labor cost arbitrage on AI driven automation is so large that even mediocre execution can win significant share before the incumbents fully wake up.
For operators inside hub services, prior authorization vendors, patient access platforms, and revenue cycle companies, the strategic question is no longer whether the orchestration layer is going to consolidate workflows that currently live across multiple vendors. It is when, and what your moat looks like once the orchestration layer is in place. For investors, the question is whether the orchestration thesis is now too consensus to underwrite at attractive entry valuations, or whether there are still adjacent categories where the same dynamics are about to play out at earlier stages. Medical benefit drugs, behavioral health, durable medical equipment, complex referrals, and any other workflow where multiple fragmented stakeholders coordinate around high cost decisions are all candidates, and the alpha will go to investors who can identify the next category before the multiples reprice.
For everyone else, the simplest summary is that healthcare’s most important infrastructure companies of the next decade probably will not look like the most important infrastructure companies of the last one. They will sit further down the operational stack, they will monetize transactions rather than seats, they will be hated by some incumbents and loved by others, and they will compound on network effects rather than feature parity. Forus is the first of these to clear the visibility threshold of a billion dollar valuation with a credible biopharma customer roster and a real product in market. It will not be the last, but the fact that it exists at all is a useful signal that the market has finally figured out where the money actually is.
