London-based Scarlet has raised $50 million in Series B funding to expand a certification platform built specifically for AI and software medical devices, as regulators and healthcare companies struggle to keep approval processes aligned with rapidly changing software.
The round was led by Headline, with participation from Notion Capital, Atomico, Creandum, Kindred Capital and Entrepreneurs First. Headline partner Nancy Xiao McLaughlin is joining Scarlet’s board.
Scarlet describes itself as a global AI certification body for medical devices, combining regulatory expertise with software designed around products that change far more frequently than traditional medical equipment.
Medical AI is moving faster than traditional certification
The problem Scarlet is targeting is increasingly structural.
A company can develop or retrain an AI diagnostic system in weeks, yet assessment by certification experts can take 12 to 18 months. Even relatively minor product changes can require another six to eight months of review, according to the company.
That timeline clashes with how software is actually developed.
Scarlet’s platform allows companies to upload technical documentation digitally before qualified reviewers assess the submission against the relevant regulations and ultimately decide whether the product can enter the market. The company emphasizes that humans still make certification decisions rather than delegating them entirely to AI.
Scarlet is designated under European and UK medical-device frameworks and recognized by the U.S. FDA as a Third Party Review Organization.
Former Babylon Health builders are tackling the regulatory layer
Scarlet was founded in 2021 by James Dewar and Jamie Cox, who previously worked on AI medical devices at Babylon Health.
The company spent its first three years working with regulators before issuing certifications, according to Tech.eu. Its customers now include Flo Health, DeepHealth, NHS England, Skin Analytics, Heidi Health and Tandem Health.
The new funding will support expansion into the United States and Asia-Pacific, additional expert hiring and broader certification capabilities.
That is important because the certification bottleneck will grow if healthcare companies deploy more frequently updated AI systems.
Regulation is becoming part of the AI infrastructure stack
Scarlet’s funding is interesting precisely because it is not another investment in a medical model.
It is investment in the infrastructure around those models.
AI healthcare companies need datasets, compute and clinical validation, but they also need a regulatory process capable of evaluating systems whose behavior and underlying software may continue changing after launch.
Scarlet’s website explicitly emphasizes the ability to ship frequent releases while working with certification designed around iterative product development.
For European founders, that suggests another layer of the AI economy is becoming venture-scale: the systems that determine whether regulated AI can move from a laboratory into production.
The faster medical AI develops, the more valuable infrastructure that can evaluate those products quickly without making safety move at software speed may become.
