Ketryx has raised $39m to support the ongoing development of its AI-powered compliance platform.
The US-based company’s platform deploys tools that automate documentation, create traceability, and accelerate release cycles for clients involved in AI-based medical device development.
Transformation Capital led the Series B funding round, which brings Ketryx’s funding to date to more than $55m. Other participants included existing investors such as Lightspeed Venture Partners, Ubiquity Ventures, and 53 Stations.
Alongside building out its development roadmap, Ketryx said the latest funding would also support the growth of teams at its site in Boston and Vienna, Austria.
While Ketryx deals with cybersecurity and privacy issues relating to medical device product development, a key focus of its platform is to assist with validation protocols that help medtech manufacturers demonstrate that their product does what they claim it does.
Ketryx’s CEO Erez Kaminski told Medical Device Network: “When people say it's hard to build a medical device, what they mean is that it’s hard to ensure the device is high-quality, has been developed correctly, and is compliant and well-documented.
“We're a compliant solution that helps companies that build medical devices to effectively click a button and generate all the evidence related to these factors in a safe way through an agentic AI workflow.”
According to GlobalData analysis, AI in healthcare is forecast to reach a $19bn valuation by 2027.
Increasingly, medtech companies want to use AI in their product, and they also want to use AI as part of their development process.
Kaminski continued: “On one hand, Ketryx is a platform that helps companies accelerate the rate of development for medical AI, and on the other, we serve as the way in which companies use AI to develop their products.”
To meet these aims, Kaminski said that Ketryx clients tend to require assistance in two respects that its platform helps to fulfil – in creating a more qualified and validated AI product, and safely utilising AI.
Kaminski added: “Part of that is ensuring that in critical decision making that can impact patient safety, a human has to be involved in that decision, and that it can be proven that the AI cannot make changes to your product without human approval.
“I don't believe you can mitigate the chance of AI making hallucinations because it's tied to what it does. AI hallucinates all the time. It's just when it's right, we call it a good answer, and when it's wrong, we call it a hallucination. But it does the same thing in both cases.
Kaminski concluded: “What I believe is that for some use cases of AI’s application in medical devices, you can statistically prove that it's good enough or better than humans. While for others, a human in the loop may still be necessary to help prove that the AI works.”
