Nelson Advisors: Accurx Scribe, powered by Tandem - European HealthTech Case Study
How a London communications platform and a Stockholm AI company built the NHS's fastest scaling ambient scribe, and what founders and investors should take from it
Introduction
Strategic partnerships in HealthTech are announced often and delivered rarely. The press release is easy; the shared roadmap, the aligned commercial incentives and the willingness to let another company sit between you and your customer are hard. Which is why Accurx Scribe, powered by Tandem, deserves attention as one of the more instructive European HealthTech stories of the last eighteen months.
In May 2025, Accurx, the London-based communications platform used by 98% of GP practices in England and more than 200,000 NHS staff, announced that it would embed Tandem Health's ambient AI scribe into its own product rather than build one from scratch. Sixteen months later, Accurx Scribe has secured 25 Trust-wide contracts, with a significant pipeline of pilots and proofs of concept expected to convert over the coming months as NHS England's Frontline Productivity funding flows through to the frontline. In March 2026, a joint four-year programme across University Hospitals of Leicester and University Hospitals of Northamptonshire put the tool in front of 10,000 clinicians covering roughly 2.5 million outpatient appointments a year. And on 14 September 2026, Tandem closed a $100 million Series B led by EQT's Scaleup Europe Fund, taking its total funding to $160 million.
This article looks at why the partnership has worked, what each side has gained, what it teaches founders and investors about partnering in regulated markets and where the risks lie as both companies push deeper into the clinical and administrative workflow of the NHS.
The Partnership in Brief
The division of labour is clean. Tandem provides the underlying ambient scribe: the speech capture, transcription, summarisation and document generation engine, plus the clinical safety evidence that sits behind it. Accurx provides the distribution, the identity layer, the integration with clinical record systems and the user interface that clinicians already have open on their desktop, web and mobile devices. Tandem builds the engine; Accurx builds the car, and the car is already parked in almost every GP surgery and a growing number of hospital departments in England.
That distribution advantage is the heart of the story. Accurx started life in 2016 as a messaging tool for GP practices and grew, largely through product-led adoption during the pandemic, into the default communication layer between NHS staff and patients. When Accurx Scribe launched in April 2025, it did not have to win a single new login. It appeared as a feature inside a product that clinicians were already using multiple times a day, with write-back to SystmOne and EMIS records and the ability to generate referral letters, Advice and Guidance requests, appointment summaries and patient follow-up messages from the same consultation capture.
Tandem, founded in Stockholm and led by CEO Lukas Saari, had been building a first party product across the Nordics and Europe. By the time of the partnership it had a growing evidence base, a clinical team of in-house doctors and a product used across thousands of care organisations. What it lacked in the UK was reach. Accurx provided reach at a scale no direct sales effort could have matched.
Why It Has Worked: The Timing of the NHS Market
Partnerships succeed or fail partly on their own merits and partly on the market they enter. Accurx and Tandem timed theirs well.
In 2025, NHS England moved from cautious guidance on ambient voice technology to active encouragement. The MHRA clarified that most AVT products generating clinical documentation are medical devices, which raised the bar for suppliers but also legitimised the category. NHS England then published the Ambient Voice Technology Self-Certified Supplier Registry, listing 19 suppliers in January 2026 who could demonstrate Class 1 medical device registration, a current DTAC assessment, evidence of benefit in NHS settings, integration capability with existing NHS infrastructure, scalability, performance monitoring and an indicative price list. Both Accurx and Tandem are on that registry, in their own right.
The registry matters because it turned a chaotic market of pilots into something Trusts could procure against with confidence. But the more important development was money. The Frontline Productivity programme, NHS England's plan to fund digital tools that release clinical time, is now beginning to flow to Trusts. Fully funded AVT pilots were opened to Trusts during 2026, and the conversion of those pilots into multi-year contracts is what will define the winners in this category over the next twelve months.
Accurx Scribe was positioned ahead of that wave rather than behind it. Twenty-five Trust-wide contracts before the bulk of central funding has arrived is a strong leading indicator. When the funding lands, procurement teams tend to gravitate towards suppliers who are already live in comparable Trusts, already integrated with the relevant EPR and already on the national registry. Accurx ticks all three boxes, and its Scribe pipeline reflects that.
Benefits for Accurx
The most obvious benefit to Accurx is speed to market. Building a clinically safe ambient scribe from nothing is not a six-month project. It requires speech models tuned to clinical vocabulary and regional accents, a clinical safety case, medical device registration, a large evaluation effort and a team of clinicians to validate outputs. Tandem had done that work. By partnering, Accurx launched a credible product in the same quarter that the market began to move, rather than a year later.
The second benefit is strategic. Accurx has always faced the question of what sits beyond messaging. A communication layer is valuable but potentially commoditised, and the company has spent years extending into triage, batch messaging, patient-initiated follow-up and, increasingly, secondary care. Ambient scribing pulls Accurx into the consultation itself, the moment where clinical value is created. Once Accurx is capturing the consultation, it is natural for it to own the documents, the letters, the coding and the follow-up communication that flow from that consultation. Scribe converts Accurx from a tool that sends messages into a platform that generates the content of those messages.
That is precisely the direction the company is now taking. Accurx is extending Scribe into the end-to-end clinical and secretarial workflow: not just capturing the consultation and generating the clinical document, but managing the correspondence that follows, including sending it digitally to patients with physical letters as a fallback. In the Leicester and Northamptonshire pilot, patient letters went out the same day rather than two to three days later, and clinicians saved around eight minutes of documentation per patient and roughly an hour of administrative time per day.
The third benefit is commercial. Accurx already integrates with several major Trust EPRs, a capability that many suppliers on the registry do not have. That integration, combined with Scribe, lets Accurx sell not merely a time-saving tool for clinicians but a replacement for parts of the legacy transcription, document and letter-management infrastructure that Trusts still pay significant sums for today. Outsourced medical transcription, hybrid mail contracts and document management middleware are large, unglamorous line items in Trust budgets. Displacing them is a bigger prize than a per-clinician scribe licence, and it is a prize that a pure-play scribe vendor without communication rails cannot easily reach.
Benefits for Tandem
For Tandem, the arithmetic is simple. The UK is one of Europe's largest healthcare markets, but selling to the NHS from Stockholm is slow, expensive and dependent on relationships that take years to build. Accurx handed Tandem a route into more than 200,000 NHS staff and, through the Trust-wide contracts now in place, tens of thousands of hospital clinicians it would have struggled to reach directly.
That scale has fed back into the product. Ambient scribes improve with volume and variety of use, and the NHS gives Tandem exposure to a healthcare system with a distinctive vocabulary, a distinctive set of documents, and an unusually diverse patient and clinician population. Every consultation captured through Accurx Scribe makes Tandem's models better tuned to English-language clinical practice, an asset that carries into other markets.
The partnership has also provided proof points that matter to investors. When Tandem raised its $50 million Series A in July 2025, led by Kinnevik with Northzone and Amino Collective participating, the Accurx partnership and NHS traction were central to the pitch. When it raised $100 million in September 2026 from EQT's Scaleup Europe Fund, with the same investors returning, the story had evolved: Tandem now describes itself as building an AI-native "clinic operating system" covering patient flow, triage, scheduling and patient communications across 14 European markets. The NHS, reached via Accurx, is one of the more visible reference deployments behind that ambition.
Finally, the partnership has allowed Tandem to concentrate its own resources where it is strongest. Rather than building a UK sales force, integrating with SystmOne and EMIS, or navigating the NHS's information governance landscape practice by practice, Tandem could invest in model quality, clinical safety, and expansion into Germany, France, Spain and the rest of Europe. The Accurx deal is, in effect, a capital-efficient UK go-to-market that let Tandem spend its Series A on product and geographic expansion instead.
An Example of a Successful Strategic Partnership
What distinguishes this from the many HealthTech partnerships that go nowhere?
The first factor is complementarity without overlap. Accurx had distribution and integration; Tandem had the engine. Neither was pretending to have the other's asset. Partnerships between two companies that both secretly want to own the whole stack tend to collapse into rivalry within a year. This one had a clear answer to "who does what" from the start.
The second is that both parties had something meaningful at stake. Accurx put its brand on the product. "Accurx Scribe, powered by Tandem" is a co-branded proposition where Accurx carries the customer relationship and the reputational risk if the scribe makes a mistake, while Tandem's name signals the technical provenance. That mutual exposure creates discipline: neither side can treat the partnership as a side project.
The third is that the partnership fits the customer's buying behaviour. NHS Trusts prefer to buy fewer things from fewer suppliers, especially where information governance, clinical safety and integration are involved. A Trust that already has an Accurx contract, an Accurx DPIA and an Accurx EPR integration can add Scribe with far less friction than it could onboard a new supplier. The partnership reduces the customer's procurement burden, which is a form of value that founders often underestimate.
The fourth is speed of iteration. Because Tandem owns the model layer and Accurx owns the workflow, both can improve their part without waiting on the other. Tandem can ship better summarisation; Accurx can ship letter-management and patient communication features. The joint product improves on two tracks at once.
Key Lessons for Founders
The most important lesson is to be honest about which asset you actually have. Accurx could have built a scribe. Its engineering team is strong, and there was internal temptation, as there always is, to own the model. It chose not to, recognising that its scarce and defensible asset was distribution and integration, not speech to text.
Founders who insist on building everything in house often lose the market to competitors who partner and ship first. Speed to market in a category that is being shaped by national policy and central funding is worth more than architectural purity.
The second lesson is that distribution is a product. Tandem's decision to go through Accurx rather than around it was not a concession; it was a recognition that the UK market rewards whoever is already on the clinician's screen. European founders expanding into the NHS should ask who already owns the workflow they want to enter and whether partnering with that company is faster and cheaper than displacing them.
The third lesson is to align with the regulatory and funding calendar. Both companies invested early in medical device registration, DTAC and clinical evidence, which put them on the AVT registry in the first cohort. That early compliance investment, which many startups see as overhead, turned into a moat when the registry became the procurement gateway. Founders selling into the NHS should treat regulatory readiness as a go-to-market activity, not a legal one.
The fourth is to sell the outcome, not the feature. Time saved per consultation is a nice metric, but Trusts do not have a budget line called "clinician time." They do have budget lines for transcription, document management and hybrid mail. Accurx has been disciplined about framing Scribe as infrastructure replacement rather than a productivity gadget, and that framing is what unlocks Trust-wide, multi-year contracts rather than departmental pilots.
The fifth lesson, and the least comfortable, is to think about what happens when the partnership succeeds. Success creates its own pressures, and both companies are now large enough that their interests will not always align perfectly. Founders should negotiate for the day the partnership works, not just the day it launches.
Key Lessons for Investors
For investors, the case study challenges a common assumption that partnerships signal weakness. Tandem's investors backed a company whose largest single market was reached through someone else's brand. That would make some venture investors nervous. Yet it is precisely because the partnership worked that Tandem could raise $150 million across two rounds in fourteen months while spending its capital on product and European expansion rather than a UK sales team.
The question investors should ask is not "does the company own the customer?" but "does the company own something the customer relationship depends on?" Tandem owns the model, the clinical safety evidence and, increasingly, the data flywheel. Accurx owns the workflow, the integrations and the trust of NHS IT and IG teams. Both are defensible. Neither is complete alone.
A second lesson is that in publicly funded health systems, the timing of central funding programmes is a leading indicator that private-market investors frequently miss. The Frontline Productivity programme, the AVT registry and the funded pilots were all visible in policy documents months before contracts appeared. Investors who track NHS England's policy calendar as closely as they track a company's ARR will price these opportunities better.
A third lesson concerns category consolidation. With 19 suppliers on the registry and a handful of them, including Microsoft Dragon, Heidi, Tortus, Corti and Lyrebird, competing for the same Trusts, the category will not support that many winners. The combination of national distribution, EPR integration and a deep-pocketed model partner is a plausible profile for one of the survivors. Investors should assess whether the companies they back have a path to that profile, on their own or through partnership.
Potential Problems and Risks for Both Companies
No case study is complete without the uncomfortable part. The same features that make this partnership strong create risks.
For Accurx, the central risk is that it does not own the model. As ambient scribing becomes a commodity, and it will, differentiation will migrate to the workflow layer, which favours Accurx. But if the quality of the underlying scribe becomes the deciding factor in a Trust procurement, Accurx is only as good as Tandem's model and Tandem's willingness to prioritise Accurx's requirements over those of its other markets and, potentially, its other channel partners. A large European company with $160 million in funding and a stated ambition to be a "clinic operating system" will not indefinitely be content to be the engine inside someone else's brand in Europe's most valuable single-payer market.
For Tandem, the mirror-image risk is that it does not own the UK customer. Accurx carries the relationship, the contract and the data flows in most Trust deployments. If Accurx were to develop or acquire its own model, or to add a second scribe provider to its platform to reduce dependency, Tandem's UK revenue could contract quickly. Tandem also faces reputational exposure it does not fully control: a clinical safety incident in an Accurx deployment would attach to Tandem's name regardless of whether the root cause lay in the model, the integration or the workflow.
Both companies face the risk of a well funded incumbent. Microsoft Dragon Copilot is on the registry and is being deployed at scale in Manchester and elsewhere, bundled with the Microsoft 365 estate that every Trust already owns. Epic, Oracle Health and other EPR vendors are building native ambient documentation into their platforms. In the long run, Trusts may prefer an EPR-native scribe to a third-party one, however good the third party's integration. The Accurx-Tandem proposition depends on staying enough ahead on workflow and evidence to justify a separate contract.
There is also policy risk. Central funding programmes are generous when they start and unreliable when they end. If Frontline Productivity money is slower or smaller than expected, or if it is diverted to other priorities, the pipeline of pilots that both companies are counting on could stall. Trusts that have used central funding for a two-year pilot may struggle to find recurring money when that funding expires, which is the classic NHS "pilotitis" problem.
And there is regulatory risk. The MHRA's clarification that most AVT products are medical devices was helpful in defining the category, but regulators tend to raise requirements as usage grows. A shift from Class I to Class IIa classification for products that generate clinical recommendations, or tighter rules on data residency and model training, could impose costs that a smaller supplier finds harder to bear than a Microsoft.
Potential Issues on the Critical Path
Beyond the strategic risks are the practical bottlenecks that will determine whether 25 Trust wide contracts become 75.
The first is EPR integration depth. Accurx's existing integrations with major Trust EPRs are a genuine advantage, but ambient scribing places new demands on those integrations. Writing a structured clinic letter, populating coded fields and routing correspondence to secretarial teams requires deeper write-back than sending a text message. Every EPR has different APIs, different governance and different appetite for third-party write access. Each new integration is a multi-month project, and Trusts will not convert pilots to contracts if the integration is shallow.
The second is the secretarial workflow. Replacing legacy transcription and letter-management infrastructure is a bigger prize than clinician time-saving, but it is also a more complex change. Medical secretaries, clinic coordinators and outsourced transcription providers all have a stake in the current process. Trusts will need change-management support, redesigned processes and clear governance for who signs off letters generated by AI. Accurx is taking on a services-heavy transformation, not just a software rollout.
The third is clinical safety at scale. Scribe outputs are reviewed by clinicians before they are filed, but as usage grows, review fatigue becomes a real issue. A model error that slips through review in one of 2.5 million outpatient appointments a year is statistically inevitable. Both companies need robust monitoring, incident reporting and a shared clinical safety framework that stands up to scrutiny from Trust clinical safety officers and, potentially, coroners.
The fourth is commercial model. Trust-wide contracts on a per-clinician basis are simple to buy but expose the supplier to underutilisation. If Trusts pay for 5,000 licences and 2,000 clinicians actually use the tool, renewal conversations become difficult. Accurx and Tandem will need to demonstrate adoption, not just contract signature, and share the revenue in a way that keeps Tandem invested in UK success.
The fifth is capacity. Both companies are scaling quickly. Accurx is moving from primary care, where it is dominant, into secondary care, where its position is newer. Tandem is expanding into 14 markets simultaneously. The risk that the UK partnership becomes a lower priority for one side while the other is counting on it is real, and it is the kind of misalignment that surfaces only when things are going well enough for both companies to have other options.
Platform Dependency
The deepest question the case study raises is one of platform dependency, in both directions. Tandem is dependent on Accurx as a channel in the UK. Accurx is dependent on Tandem as a technology supplier. Both are dependent on NHS England for the regulatory framework, the registry and the funding that is driving adoption. And both are dependent on EPR vendors for the integrations that make the product useful.
Mutual dependency is not inherently unstable. It is the basis of most successful supply relationships. But it requires contractual and commercial structures that survive the moment when one party's alternatives improve. Investors and founders looking at this partnership should ask a set of questions to which outsiders do not have the answers. How long is the exclusivity, if any, in either direction? Who owns the data generated by NHS consultations, and can it be used to train models that Tandem deploys in other markets or through other channels? What happens to a Trust-wide contract if the partnership ends? Does Accurx have a contractual right to a second model provider, and does Tandem have a right to sell directly into the NHS in competition with its own partner?
The most likely long-term evolution is one of two paths. Either the partnership deepens into something closer to a joint venture, with shared roadmaps, shared economics and possibly cross-shareholdings, or one party acquires the capability of the other. Accurx could build or buy a model; Tandem could build or buy UK distribution. Neither path is bad for the NHS, which will get a better product either way, but both paths carry execution risk and, in the acquisition case, the risk that value is transferred rather than created.
There is also a broader platform dependency worth naming. Both companies sit on top of foundation models supplied by a small number of large AI labs, and both compete, directly or indirectly, with the hyperscaler that also supplies infrastructure to most NHS Trusts. That is a dependency every ambient scribe company shares, and it is the one least within any founder's control.
Conclusion
Accurx Scribe, powered by Tandem, is a rare example of a European HealthTech partnership that has delivered measurable scale rather than a press release. In sixteen months it has produced 25 Trust-wide contracts, a 10,000-clinician multi-Trust programme, a place on the national registry for both companies and a significant contribution to Tandem's $160 million of funding. It has done so because the two companies brought complementary assets, aligned with the NHS's regulatory and funding calendar, and were disciplined about selling infrastructure replacement rather than a feature.
The lessons for founders are to be honest about which asset you own, to treat distribution and regulatory readiness as products, and to negotiate for the day the partnership succeeds. The lessons for investors are to look past who owns the customer to who owns what the customer relationship depends on, and to read the policy calendar as carefully as the metrics.
The risks are real. Neither company owns the whole stack. Both face incumbents with deeper pockets and EPR vendors with more natural integration. Central funding is generous now and uncertain later. And the mutual dependency that makes the partnership work today is the same dependency that could strain it tomorrow.
For now, though, this is what a successful strategic partnership in European HealthTech looks like: two companies, each doing what it does best, in a market that finally has the policy, the funding and the appetite to reward them for it. The next twelve months, as the pilot pipeline converts and Frontline Productivity funding lands, will show whether the model holds.
Nelson Advisors > European HealthTech, MedTech, Digital Health Investment Banking
Nelson Advisors specialise in Mergers and Acquisitions, Partnerships and Investments for Digital Health, HealthTech, MedTech, Health IT, Consumer HealthTech, Healthcare Cybersecurity, Healthcare AI companies.www.nelsonadvisors.co.uk
Nelson Advisors regularly publish Thought Leadership articles covering market insights, industry trends, deal commentary, market analysis & predictions @ https://www.healthcare.digital
Nelson Advisors publish Europe's Leading Healthcare Technology Investment Banking Newsletter every week, join 5000+ HealthTech and MedTech subscribers today! https://lnkd.in/e5hTp_xb
Nelson Advisors pride ourselves on our DNA as ‘Founders advising Founders.’ We partner with entrepreneurs, boards, corporates, venture capital and private investors to maximise shareholder value and investment returns.www.nelsonadvisors.co.uk
#NelsonAdvisors #HealthTech#MedTech#DigitalHealth #HealthIT #Cybersecurity #HealthcareAI #FemTech#ConsumerHealth #Mergers #Acquisitions #Partnerships #Growth #Strategy #NHS #UK #Europe #USA#Canada#Commonwealth#CorporateDivestitures #VentureCapital #PrivateEquity #Founders #SeriesA #SeriesB #Founders #SellSide #TechAssets #Fundraising #BuildBuyPartner #GoToMarket #PharmaTech #BioTech #Genomics
Nelson Advisors LLP
Hale House, 76-78 Portland Place, Marylebone, London, W1B 1NT
Meet Nelson Advisors @ 2026 Events
Digital Health Rewired > March 2026 > Birmingham, UK
NHS ConfedExpo > June 2026 > Manchester, UK
HLTH Europe > June 2026, Amsterdam, Netherlands
HIMSS AI in Healthcare > July 2026, New York, USA
Bits & Pretzels > September 2026, Munich, Germany
World Health Summit 2026 > October 2026, Berlin, Germany
HealthInvestor Healthcare Summit > October 2026, London, UK
HLTH USA 2026 > October 2026, USA
Barclays Health Elevate > October 2026, London, UK
Web Summit 2026 > November 2026, Lisbon, Portugal
MEDICA 2026 > November 2026, Düsseldorf, Germany
Venture Capital World Summit > December 2026 Toronto, Canada
Comments