Dr Kola Tytler (MBBS MBA MRCGP) | 15 July 2026 | 11 min read
Neko Health's $700 million raise should be read alongside a broader pattern across medical AI: companies that started with a narrow, single-purpose function are steadily expanding into fuller, more vertically integrated systems, and investors are increasingly rewarding that expansion over the standalone feature it began as.
Where these companies started
Most of the current generation of well-known medical AI companies began narrowly. Ambient scribes produced consultation notes and little else. Clinical search platforms answered discrete clinical questions. Imaging algorithms identified specific, narrow categories of abnormality. Laboratory platforms simply displayed test results back to the patient or clinician. Each of these was a genuinely useful point solution, and each was, in principle, a feature that a larger incumbent platform, an EPR vendor for example, could eventually build or acquire.
Where they are expanding to now
The direction of travel across nearly every category has been the same: outward, from the original narrow function towards a broader slice of the care pathway. Heidi increasingly presents itself as an AI care partner spanning multiple stages of the clinical encounter rather than a documentation tool alone. Tandem has moved from scribing towards a genuinely integrated clinical workflow platform, adding a Class IIa-certified coding assistant and clinical decision support alongside its original scribe. TORTUS has built regulated, safety-focused infrastructure around ambient clinical AI rather than treating regulation as an afterthought. Function Health combined its original laboratory testing membership with Ezra's AI-enhanced imaging capability, moving from blood panels into full-body scanning under one brand. Neko, the furthest along this particular axis, combines proprietary diagnostic hardware, software, physical clinics and structured clinical follow-up into a single vertically integrated healthcare provider rather than a tool that plugs into someone else's.
Why companies choose to expand horizontally rather than stay narrow
There are consistent commercial logics behind this pattern across very different companies. A broader slice of the care pathway captures more of the total value being created, rather than ceding the adjacent, often more lucrative steps to a different vendor. It improves customer retention, since a platform embedded across multiple stages of care is harder to replace piecemeal than a single-purpose tool. It generates more proprietary data, which, as the category matures, is increasingly the more durable competitive asset relative to model capability alone. It reduces dependence on being a feature inside someone else's platform, a structurally weaker position for any standalone AI company to be in long term. And it allows a company to demonstrate clearer evidence of clinical and economic value across a full pathway, which is considerably more persuasive to large institutional buyers, NHS trusts among them, than evidence limited to one narrow step.
Why vertical integration is genuinely difficult, and not the right strategy for everyone
The same expansion that strengthens a company's competitive position also multiplies its risk and cost. Greater regulatory exposure follows naturally from touching more of the clinical pathway, as Tandem's and TORTUS's respective moves into Class IIa certified territory illustrate directly. Clinical liability broadens with scope. Operational complexity increases sharply once a company is running physical clinics, employing clinical staff, or managing hardware manufacturing, none of which resembles running a software product. Expansion of this kind is also inherently slower to scale geographically than a pure software deployment, since it typically requires local regulatory clearance, physical premises or clinical staffing in each new market, a constraint Neko will test directly as it opens its first US clinics this year.
The investor logic behind financing this kind of expansion
A full-stack healthcare provider is, in principle, more defensible against both foundation-model commoditisation and large incumbent competition than a thin AI interface sitting on top of someone else's data and someone else's customer relationship. Neko's round is a clear signal that investors are willing to finance genuinely capital-intensive clinical infrastructure, hardware, clinics, clinical staff and all, when the underlying company can demonstrate a credible AI and proprietary data advantage layered on top of it. That is a materially different, and more expensive, bet than backing a software feature, and it is one investors appear increasingly willing to make across several categories at once, not only in preventive health.
What this suggests for healthtech strategy more broadly
The clear implication for founders and platforms across the category, including in adjacent fields such as clinical education and exam preparation, is that a thin AI interface alone is unlikely to remain a defensible position for long, as underlying model capability continues to commoditise. Durable advantage looks more likely to come from some combination of proprietary, hard-to-reproduce data, genuine clinical trust built over time, workflow integration deep enough to be difficult to remove, credible evidence of real-world value, and, in some categories, direct control over data generation itself rather than only its interpretation.
Conclusion
The overall direction of travel across medical AI, from Neko's clinics to Tandem's expanding certified platform to Function's combined lab-and-imaging model, is consistently the same: away from single-purpose AI tools and towards fuller AI-enabled healthcare organisations. Not every company should, or will, attempt that transition successfully. But the ones that do appear to be attracting a disproportionate share of the capital now flowing into the category.