adsilico

Generative AI and physics-based simulation for in-silico testing of implantable medical devices.

Cover Block

Public sources

Field	Detail
Name	adsilico
Tagline	Generative AI and physics-based simulation for in-silico testing of implantable medical devices [adsilico, retrieved 2024]
Headquarters	Leeds, United Kingdom [Crunchbase, retrieved 2024]
Founded	2024 [Crunchbase, retrieved 2024]
Stage	Series A [The Yorkshire Post, October 2026]
Business model	B2B
Industry	Healthtech
Technology	AI / Machine Learning
Geography	Western Europe
Growth profile	Venture Scale
Founding team	Academic Spinout [University of Manchester, February 2024]
Funding label	Series A
Total disclosed funding	~$17,000,000 [University of Manchester, February 2024] [The Yorkshire Post, October 2026]

Executive Summary

PUBLIC adsilico builds in-silico testing software for implantable medical devices, and it merits investor attention now because public reporting shows it has moved from a 2024 seed financing to a 2026 Series A around a thesis that sits at the intersection of AI, simulation, and regulatory pressure to improve trial efficiency and diversity [University of Manchester, February 2024] [The Yorkshire Post, October 2026] [Tech.eu, October 2026]. The company is described as a University of Leeds spinout, with founding ties that also extend into the University of Manchester, which gives the story a recognizable academic-commercial pattern rather than a purely software-first startup origin [University of Manchester, February 2024] [The Yorkshire Post, October 2026].

The product case is straightforward in concept, even if technical validation remains the central diligence question: adsilico says it uses anonymised clinical scans, generative AI, and computational modelling to create virtual patient populations so device makers can test safety and performance before, and alongside, animal and human studies [adsilico, retrieved 2024] [University of Manchester, February 2024]. Public sources point first to heart valves, stents, and neurovascular devices, with 2026 coverage indicating planned expansion from cardiac work into brain devices such as flow diverters for aneurysm treatment [The Yorkshire Post, October 2026].

The founding team is relevant to the problem set. Sheena Macpherson is identified as co-founder and CEO, Dr Nishant Ravikumar as co-founder and Head of Virtual Populations, Dr Zeike Taylor as co-founder and Head of Simulations, and Professor Alejandro Frangi as a founder and Bicentennial Turing Chair in Computational Medicine at the University of Manchester [adsilico, retrieved 2024] [Northern Gritstone, retrieved 2026] [University of Manchester, February 2024] [cistib.org, retrieved 2026]. That mix suggests domain depth in computational medicine and simulation, which is necessary in this category, though investors will still want evidence that scientific credibility is translating into repeatable commercial adoption.

On financing, the disclosed path is modest but coherent: a £3.5 million seed round announced in February 2024 from Northern Gritstone and Parkwalk Advisors, followed by a £10 million Series A reported in October 2026 led by PXN Ventures, bringing total equity funding to £13.5 million [University of Manchester, February 2024] [The Yorkshire Post, October 2026] [Tech.eu, October 2026]. The business model appears B2B, selling to medical device manufacturers and developers that need earlier evidence on device performance and safety [University of Manchester, February 2024].

Over the next 12 to 18 months, the main watchpoints are whether adsilico can convert technical promise into named commercial proof points, broaden beyond initial cardiovascular applications without diluting focus, and show that its validation approach can support real-world regulatory and development workflows rather than remaining a compelling R&D tool [adsilico, retrieved 2024] [The Yorkshire Post, October 2026]. The company is addressing a genuine pain point in implantable-device development, but the public record still leans more heavily on product intent and funding momentum than on disclosed customer, regulatory, or deployment evidence [University of Manchester, February 2024] [The Yorkshire Post, October 2026].

Lightly corroborated -- Relies on independent funding coverage from the University of Manchester, The Yorkshire Post, and Tech.eu, with product-specific claims partly sourced from the company website.

Taxonomy Snapshot

Axis	Value
Stage	Series A
Business Model	B2B
Industry / Vertical	Healthtech
Technology Type	AI / Machine Learning
Geography	Western Europe
Growth Profile	Venture Scale
Founding Team	Academic Spinout
Funding	Total disclosed approximately $17,000,000 [University of Manchester, February 2024] [The Yorkshire Post, October 2026]

How the Company Got Here

PUBLIC adsilico appears to have formed in 2024 as a Leeds-based company focused on in-silico testing for implantable medical devices, with the company website describing a platform that combines generative AI and physics-based simulation for virtual patient modelling [adsilico, retrieved 2024] [Crunchbase, retrieved 2024]. Public company materials identify Sheena Macpherson, Dr Nishant Ravikumar, Dr Zeike Taylor, and Professor Alejandro Frangi among the founding team, and position the business as a University of Leeds spinout with an academic-origin profile rather than a conventional software startup from day one [adsilico, retrieved 2024] [Crunchbase, retrieved 2024].

The early chronology is narrow but reasonably clear from company-controlled sources. Crunchbase lists adsilico as headquartered in Leeds, United Kingdom, while the company website shows a 2024 operating team that included Macpherson as CEO, Ravikumar as Head of Virtual Populations, Taylor as Head of Simulations, and Anthony Holmes as Chief Product Officer [Crunchbase, retrieved 2024] [adsilico, retrieved 2024]. By 2024, the company was already presenting a defined thesis: use anonymised clinical scans, computational modelling, and simulation to support preclinical evidence generation for implantable devices, with validation against real patient data under clinical oversight and model credibility assessed against ASME V&V 40 principles, according to the website [adsilico, retrieved 2024].

Lightly corroborated -- Based on company website and Crunchbase, with core company identity corroborated across both but several operating details still sourced primarily from company-controlled materials.

Product and Technology

Sources and analysis

adsilico is making a specific bet: that implantable-device developers will pay for earlier, more representative performance evidence before they commit as heavily to animal studies and human trials. Public materials describe a platform that combines anonymised clinical scans, generative AI, and computational modelling to create virtual patient populations for in-silico testing of implantable medical devices [adsilico, retrieved 2024] [University of Manchester, February 2024]. The company frames the product around diversity across patients rather than a single canonical model, with initial work focused on virtual hearts and blood vessels, and with intended use in device-safety and performance assessment before, and alongside, conventional testing pathways [adsilico, retrieved 2024] [University of Manchester, February 2024] [Tech.eu, October 2026].

The product claims that matter most are methodological rather than cosmetic. On its website, adsilico says its virtual populations are validated against real patient data under clinical oversight and that model credibility is assessed using the risk-informed principles of ASME V&V 40, a standard often referenced in computational modeling for medical applications [adsilico, retrieved 2024]. Those are company-supplied claims, so they should be read as a statement of intended validation discipline rather than independent proof of regulatory acceptance or clinical adoption. Press coverage adds some scope around use cases, reporting initial applications in heart valves, stents, and neurovascular devices, with a publicly announced plan to extend cardiac work into brain devices such as flow diverters for aneurysm treatment [The Yorkshire Post, October 2026].

A small hiring signal suggests the technical stack likely includes applied machine learning dedicated to virtual-patient generation, although the public evidence remains narrow. A LinkedIn posting captured for a Machine Learning Scientist focused on virtual patient populations supports the view that ML capability is central to the product build [LinkedIn, retrieved 2026]. Beyond that, claims about infrastructure, deployment model, or regulatory workflow would be inference rather than reporting, and the public record reviewed here does not support going further.

Lightly corroborated -- Core product positioning is corroborated by the company website and University of Manchester coverage, while validation and technical-method claims remain primarily company-sourced.

Where the Demand Sits

PUBLIC This market matters now because implantable-device developers face a measurable pressure point: clinical evidence is getting more expensive and harder to generalize across diverse patient populations, while regulators and developers are showing more interest in simulation as a complement to animal and human studies [University of Manchester, February 2024] [The Yorkshire Post, October 2026] [Tech.eu, October 2026].

adsilico is not entering a neatly bounded software category with a disclosed third-party TAM in the source set. The public evidence supports a narrower description: the company is aimed at manufacturers and developers of implantable medical devices that need earlier evidence on safety and performance, particularly for products such as heart valves, stents, and neurovascular devices [University of Manchester, February 2024] [The Yorkshire Post, October 2026]. That points to a market sitting at the intersection of medtech R&D software, computational modeling, and preclinical evidence generation, rather than a generic AI tooling budget [adsilico, retrieved 2024] [Tech Funding News, October 2026].

The demand signal in the cited research is less about top-line market size and more about cost and process friction. The University of Manchester's 2024 funding announcement frames the problem as high development cost, long timelines, limited patient diversity in studies, and continued dependence on animal testing for implantable-device programs [University of Manchester, February 2024]. Later coverage in 2026 keeps that framing intact and adds a practical use case: generating virtual patient populations from anonymised scans and computational models so device makers can test performance before, and alongside, conventional studies [The Yorkshire Post, October 2026] [Tech.eu, October 2026].

That matters because the addressable demand is likely to be pulled first by programs where anatomy and hemodynamics are central to device performance. Public reporting places adsilico's initial work in heart valves, stents, and neurovascular devices, with an explicit plan to extend from cardiac applications into brain devices such as flow diverters for aneurysm treatment [The Yorkshire Post, October 2026]. In practice, that suggests the early commercial wedge is not all medtech, but a subset of implantables where virtual populations could change design iteration, preclinical testing strategy, or trial planning if regulators and customers accept the evidence package [adsilico, retrieved 2024] [The Yorkshire Post, October 2026].

Adjacent markets are easier to identify than direct market-size figures. One adjacent market is broader simulation and computational medicine software used in device design and performance analysis; another is the clinical-trial-enablement layer that helps sponsors think about cohort diversity and study feasibility, even if adsilico's current positioning is device-focused rather than trial-operations-focused [University of Manchester, February 2024] [Tech.eu, October 2026]. The closest substitute is still the incumbent preclinical pathway itself: bench testing, animal work, and human studies. Public sources describe adsilico's approach as an augmentation to that pathway, not a replacement for clinical evidence, which implies adoption risk but also a clearer fit with existing procurement and regulatory behavior [University of Manchester, February 2024].

Regulatory and macro forces in the source set point in the same direction, though the evidence remains early. The company says its model credibility is assessed using the risk-informed principles of ASME V&V 40, a framework used for credibility assessment in computational modeling for medical applications [adsilico, retrieved 2024]. That does not establish regulatory acceptance by itself, but it is consistent with a market environment in which developers need more structured validation standards if simulation outputs are to influence design or evidence generation. At the same time, the broader macro argument is straightforward: any tool that can compress R&D timelines, widen represented patient anatomy, or reduce animal-testing burden has a plausible budget owner in medtech, provided the outputs are credible enough to survive internal quality and regulatory review [University of Manchester, February 2024] [Tech Funding News, October 2026].

Market lens	Public evidence	Implication for adsilico
Core served market	Implantable medical-device developers, initially in heart valves, stents, and neurovascular devices [University of Manchester, February 2024] [The Yorkshire Post, October 2026]	Focused initial wedge rather than broad horizontal software spend
Demand driver	High cost, long timelines, limited patient diversity, and animal-testing burden in device development [University of Manchester, February 2024]	Clear ROI narrative if simulations shorten iteration cycles or improve study design
Adjacent market	Computational medicine and simulation-led evidence generation for medtech [adsilico, retrieved 2024] [Tech.eu, October 2026]	Expansion path could extend beyond one device class if validation holds
Substitute workflow	Bench testing, animal studies, and human trials remain the incumbent path [University of Manchester, February 2024]	Adoption depends on augmenting existing evidence stacks, not displacing them outright

The table shows a market defined more by workflow pain than by a published category total. For investors, that usually means the near-term question is not headline TAM, but whether the company can win repeat budgets inside a tightly specified medtech R&D motion before it broadens into a larger simulation platform.

Lightly corroborated -- Relies primarily on University of Manchester, The Yorkshire Post, Tech.eu, and the company website; market framing is supported, but no independent third-party TAM study appears in the cited public sources.

Competitive Landscape

MIXED adsilico appears to sit in a narrow but consequential part of medtech tooling, closer to a computational evidence layer for implantable-device developers than to a general-purpose clinical AI vendor, but the public record here does not name direct competitors with enough precision to support a like-for-like table [University of Manchester, February 2024] [adsilico, retrieved 2024].

The practical competitive map is therefore best read by substitute set rather than by logo. At one end are incumbent device-development workflows built around bench testing, animal studies, and later-stage human trials, which remain the default evidence path for implantable devices and the burden adsilico is explicitly trying to reduce or augment rather than replace [University of Manchester, February 2024]. Adjacent to that are specialist simulation and modelling approaches used earlier in R&D, where adsilico's pitch is that generative AI plus computational modelling can create more diverse virtual patient populations than a conventional engineering-only workflow [adsilico, retrieved 2024] [Tech Funding News, October 2026]. A third substitute set sits outside medtech altogether, namely broader AI vendors selling model development, analytics, or workflow software into healthcare, but those groups are not described in the cited sources as offering implantable-device-specific in-silico testing [Tech.eu, October 2026].

The company's clearest edge today is domain specificity tied to academic origin. Public sources consistently place the business as a University of Leeds spinout with founders linked to Leeds and Manchester, including Professor Alejandro Frangi in computational medicine and Zeike Taylor in mechanical engineering, which suggests a talent base built around the intersection of imaging, simulation, and implantable-device physics rather than a generic software stack [University of Manchester, February 2024] [cistib.org, retrieved 2026] [alphaXiv, retrieved 2026]. That matters because implantable-device validation tends to reward technical credibility and clinical framing, not only model performance. The edge is real, but it is also perishable if it remains concentrated in founding expertise and company-asserted validation claims such as ASME V&V 40 alignment and clinical oversight, which are currently sourced primarily to the company itself [adsilico, retrieved 2024].

Exposure follows from the same setup. adsilico's public moat appears to rest on know-how, early investor backing, and a specific problem definition, yet no named commercial customers, formal device-company partnerships, or regulatory clearances were identified in the reviewed sources [University of Manchester, February 2024] [The Yorkshire Post, October 2026]. That leaves the company exposed to better-capitalized simulation incumbents or device manufacturers building comparable capabilities internally, especially if procurement teams prefer established vendors with deeper regulatory track records. It is also exposed to adjacent AI companies if the value shifts from high-fidelity simulation toward workflow integration or evidence packaging, two channels that the public materials do not show adsilico controlling [adsilico, retrieved 2024] [Tech.eu, October 2026].

Over the next 18 months, the most plausible competitive scenario is a split market in which specialist in-silico vendors win discrete programs while incumbents retain system-of-record status. If adsilico can convert its cardiac and neurovascular focus into named manufacturer relationships or repeat project work, it is the most credible public candidate in this source set to win share in virtual testing for implantables because the company is already framed around that use case and has raised a cumulative £13.5 million in equity funding to expand it [The Yorkshire Post, October 2026] [Tech.eu, October 2026]. If, instead, buyers continue to treat virtual populations as an experimental add-on rather than a procurement line item, the likely loser is adsilico itself, not because the technical premise is weak, but because a specialized spinout without disclosed customer references has less room for commercial delay than larger substitute categories do [University of Manchester, February 2024] [The Yorkshire Post, October 2026].

Lightly corroborated -- Core positioning, funding, and founder-background claims are supported by public sources including the University of Manchester, The Yorkshire Post, Tech.eu, and university profiles, but the absence of named direct competitors or public customer references limits corroboration for the competitive map.

Opportunity

Upside case

PUBLIC The prize here is not a point solution for simulation, but a chance to become part of the evidence-generation stack used by implantable-device makers before and alongside animal and human trials, in a market where development cost, timeline, and patient-representation constraints are already well documented [University of Manchester, February 2024].

The headline opportunity is straightforward: if adsilico can show that virtual patient populations are credible enough to shape preclinical and early clinical decisions across multiple implant categories, it could become a default computational layer for a meaningful slice of implantable-device development [adsilico, retrieved 2024] [University of Manchester, February 2024] [The Yorkshire Post, October 2026]. That is a larger outcome than selling isolated services work. The public evidence supports reachability, not certainty. The company was funded at seed in February 2024 to build product and team [University of Manchester, February 2024], then raised a £10 million Series A in October 2026 led by PXN Ventures, with prior investors Northern Gritstone and Parkwalk participating [The Yorkshire Post, October 2026] [Tech.eu, October 2026]. That financing progression suggests outside investors see a platform-shaped opportunity, while the product surface has already broadened from cardiac applications into stents, neurovascular devices, and planned work on brain devices such as aneurysm flow diverters [The Yorkshire Post, October 2026].

A second reason the upside is real is that the problem is structural rather than cyclical. adsilico is targeting high-cost, slow, and diversity-constrained development processes in implantable devices, with an approach intended to accelerate R&D, augment human trials, and reduce animal experimentation rather than claim to replace clinical evidence outright [University of Manchester, February 2024]. That positioning matters. Companies that fit into existing validation and regulatory workflows tend to have a clearer adoption path than companies that ask the sector to rewrite those workflows from first principles. The company also says its virtual populations are validated against real patient data under clinical oversight and that model credibility is assessed using ASME V&V 40 risk-informed principles, which at minimum shows it is speaking the language of regulated engineering practice [adsilico, retrieved 2024].

Scenario	What happens	Catalyst	Why it's plausible
Cardiac wedge to multi-device platform	adsilico starts with heart valves and vascular applications, then extends the same virtual-patient and simulation stack into adjacent implant categories	A successful expansion from cardiac work into neurovascular devices and brain applications such as flow diverters	Public sources already place the company in heart valves, stents, and neurovascular devices, with expansion into brain-device work explicitly described [The Yorkshire Post, October 2026]
Standard evidence layer for medtech R&D	Device makers begin treating in-silico testing as a routine part of preclinical evidence generation before animal and human studies	Repeated manufacturer adoption and validation work tied to safer, faster, and more diverse testing workflows	The company is built around preclinical assessment for manufacturers, and its stated framework emphasizes validated virtual populations and ASME V&V 40-style credibility assessment [adsilico, retrieved 2024] [University of Manchester, February 2024]
Spinout to category infrastructure	adsilico becomes a core platform used across the UK and European implantable-device ecosystem, supported by academic credibility and venture backing	Continued investor support after the seed, plus conversion of university-origin science into broader commercial programs	The business is a University of Leeds spinout with founders linked to Leeds and Manchester, and it has moved from a £3.5 million seed to a £10 million Series A in roughly two and a half years [University of Manchester, February 2024] [The Yorkshire Post, October 2026] [Tech.eu, October 2026]

The compounding mechanism would come from reuse, not consumer-style network effects. If one device program produces a validated workflow around anonymised clinical scans, virtual patient generation, and simulation, the next program should begin with more technical priors, more internal credibility, and a stronger case for adjacent use [adsilico, retrieved 2024]. Over time, that can create a data-and-workflow moat: each additional anatomy, device class, and manufacturer program could improve model calibration, shorten deployment cycles, and make adsilico harder to displace inside customer development processes. Public evidence of this flywheel is still early, but the product has already expanded beyond a single device type, and the company describes completed work across virtual hearts and blood vessels with ambitions into further neurovascular categories [adsilico, retrieved 2024] [The Yorkshire Post, October 2026] [Tech Funding News, October 2026].

The size of the win is best framed through strategic value rather than a clean public comp, because no named direct competitor or public peer is established in the source set. Even so, the company has disclosed total equity funding of £13.5 million after its Series A, which is modest relative to the clinical and regulatory importance of the workflows it is trying to influence [The Yorkshire Post, October 2026]. If the "standard evidence layer for medtech R&D" scenario above plays out, the business could justify infrastructure-like strategic value as a software-and-science platform embedded in high-value device programs, rather than as a narrow consultancy (scenario, not a forecast). The public case for that upside rests on three dated facts: a problem set with industry-level pain points, a technology stack designed to fit regulated development, and investor willingness to fund expansion from seed to Series A across more than one implant category [University of Manchester, February 2024] [adsilico, retrieved 2024] [The Yorkshire Post, October 2026].

Lightly corroborated -- This section relies on one independent university source, multiple 2026 news reports on financing and product scope, and several company-site claims about validation and technical method that remain company-sourced.

Sources