Rhys was six months old when she was diagnosed with spinal muscular atrophy, a rare disease that causes nerve cells and muscles to die over time. In the past, most infants diagnosed with SMA did not live past two. But Rhys was lucky. She was born less than a year after Zolgensma was released in 2019. With Zolgensma, survival rates have jumped from 25% to a whopping 91%. What’s more is that this wonder drug doesn’t just help patients survive, but also gives them the chance to actually live a more independent life.
But there’s a catch. In return for life-changing therapy, a single dose of Zolgensma costs over $2 million. As a result, insurers heavily scrutinize Zolgensma treatments, and patients like Rhys often wait months for approval, even though getting treatment before the six-month mark reduces your long-term recurrence risk by over 7x.
This story isn’t unique to Rhys or Zolgensma. Behind each new miracle drug is a broken infrastructure for delivering it.
Prescription medication is the fastest-growing segment of healthcare spending, which now tops over $700 billion annually in the US alone. This can be split into two broad categories: Retail and Specialty medication. Retail medications are those typically found at local pharmacies, while most advanced medications, including biologics and gene therapies like Zolgensma, fall into the latter category. You won’t find these at regular pharmacies, and there are dedicated specialty pharmacies that provide these advanced medications. Over the last decade, specialty medications have grown from just 10% to over 60% of all medication spending.
Source: Bourne Partners
The underlying process of prescribing and fulfilling medication is known as medication access management. This encompasses revenue cycle management (RCM), but also includes patient onboarding and engagement. But what’s crazy is that we’ve automated almost everything except for the 60%+ that matters most. Electronic prescribing, prior authorization, and even real-time claims adjudication are common in retail, but virtually nonexistent in specialty.
Who’s involved
The specialty medication value chain is largely separate from that of healthcare services. A typical medication care journey looks like this:
Behind the scenes, there’s a complex administrative layer that accounts for up to 40% of specialty drug spend:
![]()
Source: Avalere Health
What’s interesting here is that, unlike in healthcare services, where the physician owns the patient relationship, in specialty medication, it’s the pharmacy that typically manages patient communication and engagement. This means at scale, they (and their tech vendors) can build a patient medication dataset that is valuable to the pharmaceutical industry for both drug development and clinical trial recruitment.
Several market tailwinds are at play here. The administrative burden around medications is exploding. Now that drug spend has become the largest cost bucket, payors are tightening coverage criteria and increasing claim denials. Scrutiny over drug pricing is also intensifying across the political spectrum. Adding to the complexity, coverage policies differ significantly across PBMs and health plans, while drug-specific rebates and 340B programs, which offer discounted pricing to low-income populations, introduce additional layers of administrative and audit requirements. Operationally, specialty medications require more advanced inventory management, supply chain coordination, and clinical oversight.
At the same time, many health systems are realizing they’ve been leaving significant profit on the table by outsourcing medication fulfillment. Infusion therapies, in particular, offer attractive economics: providers can bill for both the drug and the administration, and treatment regimens are often recurring or even lifelong. As a result, systems are now investing aggressively in in-house pharmacies, infusion centers, and the supporting tech stack to capture this opportunity. This is causing huge market pull for a new tech stack.
However, the rapid growth in medication spend – and the fact that its value chain operates separately from traditional healthcare services – means that incumbent BPO and RCM providers like R1 and Ensemble have limited presence in this segment. In other words, there’s an opportunity to build an end-to-end administration platform for specialty pharmacy.
A wave of AI-native players has emerged to tackle this market. Their thesis is similar to that of healthcare services RCM. If you can capture enough of the admin process, you can unlock a take-rate business model.
Based on our conversations with current and prospective customers, we believe a 5-7% take rate is realistic today, but this could compress closer to 3% as the market becomes more competitive. Each player has a different strategy in terms of wedge workflow and market, though we see three broad categories:
Mandolin, Latent, and Plenful primarily target the pharmacy (or providers launching pharmacy services):
On the other hand, Develop Health started in digital health, where providers are often also pharmacies, and focuses on prior authorization and eligibility verification, which in turn reduces consumer drop-off in the funnel.
Lastly, Tandem Health sells to pharmaceutical companies, who then push Tandem to the prescribing providers. Pharma is willing to do this because programs like Tandem can improve time-to-medication, increase adherence (resulting in longer care journeys), and collect patient data that supports clinical trials and reimbursement negotiations with payers.
![]()
However, what really stands out compared to what we’ve seen in healthcare services is that the adoption of end-to-end solutions is happening much faster. We haven’t seen many AI RCM players in healthcare services consistently capture end-to-end billing across specialties. In most cases, vendors are seeing strong traction in one or two parts of the cycle, most commonly prior authorization and coding/chart review, with the remaining components still handled by human billers.
From what we’ve seen, prior authorization is also the hair-on-fire problem (and therefore most compelling wedge) in medication. However, we’re seeing much more appetite from buyers to adopt end-to-end solutions. We’ve heard multiple examples of large systems paying seven figures for a solution that spans the entire revenue cycle, and even some that include patient engagement:
“We adopted the platform after an RFP in 2023 and implemented them in 2024. They handle everything from prior auth, patient communications and follow-ups, 340b documentation, and claims. Since then, we reduced our staff from 35 to 3 FTEs, processing time has declined from 7 days to 24 hours, and our denial rate has dropped from 9.5% to 0.3% last quarter.”
— $2m annual spend customer
We at Northzone are excited to back the category-defining AI-native medication access management platform. In the future, patients will receive a text within minutes of their prescription being submitted, confirming insurance approval, out-of-pocket costs, and the time and location of their first dose, eliminating the days or even weeks of manual coordination that delay treatment. They’ll also get instant answers to their questions, proactive support throughout their treatment, and automated reminders to stay on track, improving adherence without creating additional work for providers.
Under the hood, the platform will orchestrate benefits verification, prior authorization, financial assistance, and pharmacy fulfillment through a network of LLM agents that parse unstructured documents, complete forms, and route tasks across the value chain in real time, regardless of the modality each stakeholder prefers. And by sitting at the center of this high-friction, high-value workflow, the platform becomes the connective tissue across stakeholders, unlocking new monetization opportunities with providers, manufacturers, pharmacies, and payers alike.
So far, we’ve mostly seen players approach this market with a software business model. It’s a tried-and-true model, and one that investors are most familiar with. That said, we’re also interested in business models that give away the software and more directly capture drug spend. For example, we’ve seen a wave of tech-enabled pharmacies, such as Alto and Pillpack, that focus on retail drugs, and we could also see an AI-native specialty pharmacy emerging in the future.
Taking a step back, this is one of the most critical pieces of the puzzle for reducing US healthcare spending. In healthcare services, RCM helps trim the fat, but that’s arguably all it does. Here, new medications are expensive, but they represent the bleeding edge of medical technology, and access truly matters – getting on a medication two weeks earlier can make a world of difference in the clinical outcome. This AI platform will play a central role in accelerating access while also building the dataset required to ensure that everyone who can benefit from these drugs can also access them.
If you’re building something in this space, we’d love to get in touch at aaron[@]northzone.com, molly[@]northzone.com, wendy[@]northzone.com!