Review and Recording: Second KIMed Webinar

“Lessons Learned from the Development and Authorisation of an AI-Based Medical Device”

News | 22 October, 2026

On October 6, 2026, the webinar “Lessons Learned – Development and Authorisation of an AI-Based Medical Device” took place as part of the “KIMed Network Impulses” webinar series. Fifty-five participants took the opportunity to gain practical insights into the development and approval of an AI-based medical device. Florian Bontrup, Managing Director of DOCYET GmbH, shared his project experience and openly discussed challenges and insights.

The Case Study: Chatbot for Patients

The case study featured the “Lotsen” chatbot for patients, which was developed by DOCYET GmbH. The system records symptoms, assesses them, and then directs users to an appropriate care option. It has a modular structure, with individual components classified as medical devices. The main product is the initial assessment.

Lessons Learned from Development and Approval

The webinar focused on specific experiences from the approval of components of the Lotsen chatbot as medical devices. Among other things, participants took away the following lessons learned:

• Conformity during development: Documentation is crucial and should accompany development from the very beginning, rather than being created retroactively.
• The correct medical device class: The choice of class has far-reaching consequences for the entire approval process and is therefore a critical decision.
• Incorporating regulatory considerations into product design: Regulatory requirements should be integrated into the product’s design from the very beginning.

Outlook: Operation and the EU AI Act

Another key focus was the EU AI Act and its implications for AI-based medical devices. Starting in 2028, additional obligations will apply to high-risk AI systems, which include certain AI systems in the medical device sector. New requirements pertain, among other things, to data governance, logging, and monitoring.

Florian Bontrup suggested viewing the AI Act as an opportunity for development: By integrating these requirements with existing regulatory standards, conformity assessment and documentation can be effectively consolidated.

In the ensuing discussion, participants exchanged views on the specific changes the AI Act will bring, the extent to which AI agents can support quality management, and the conditions under which the development of medical devices is worthwhile, even in an academic setting.

We would like to thank Florian Bontrup for his candid insights and all participants for the lively discussion!

We are happy to provide you with the recording of the webinar (in German):