Artificial Intelligence (AI) is transforming healthcare by enabling tools that act as co-pilots in everyday medical practice. This article shares the journey of developing a CE-certified AI-powered solution, and reflects on innovation, regulation, and collaboration.
In recent years, artificial intelligence (AI) has emerged as a transformative force in healthcare. From diagnostics and personalized medicine to workflow optimization, AI technologies are being embedded into clinical processes with the promise of improving patient outcomes and supporting healthcare professionals in increasingly complex medical environments. One of the most promising applications is an AI-powered assistant developed specifically for medical professionals.
This article explores the journey of creating one of the first CE-certified, AI solutions in Europe. Developed as a speech-based assistant, it supports physicians and medical staff in real-time, offering validated information at the point of care. The innovation process, key challenges, and the critical role of public funding and interdisciplinary networks in bringing this vision to life are explored.
Knowledge Healthcare professionals today are confronted with a rapidly expanding body of medical knowledge. Scientific discoveries, clinical guidelines, pharmaceutical updates, and new treatment modalities are published at a rate that far exceeds what any individual practitioner can realistically absorb. While evidence-based guidelines and institutional protocols provide structure, the daily clinical routine often demands fast, situation-specific, and accurate decision-making support—especially in time-critical or high-pressure scenarios.
The idea for this AI assistant was born from the urgent need to reduce cognitive overload while enhancing clinical precision. Instead of requiring medical professionals to search databases, consult printed manuals, or navigate complex software systems during patient care, the assistant offers a more natural, intuitive solution. Through commands, clinicians can simply ask a question—whether about drug interactions, diagnostic algorithms, or treatment pathways—and receive immediate, validated answers grounded in current medical knowledge.
From the beginning, the vision was clear: the tool must not only deliver high-quality content, but also integrate seamlessly into clinical workflows. It had to be fast, reliable, easy to use, and designed with a deep understanding of the unique demands of healthcare environments.
These ambitious goals presented a set of interrelated challenges. The assistant had to process natural language input, deliver medically accurate and legally compliant information, and function in noisy, dynamic environments such as emergency rooms and intensive care units. At the same time, it had to gain the trust of its users—physicians and nurses who bear significant responsibility and who are rightly cautious about relying on automated tools for clinical decision-making.
Therefore, usability and trust were not secondary considerations but foundational design principles. The team focused intensively on user experience, designing the voice interface to be responsive, context-aware, and capable of handling the nuanced phrasing often used in clinical language. Equally important was the establishment of transparent mechanisms for how the AI derives its answers—an aspect critical to building trust among healthcare professionals.
The development of this assistant went far beyond technological innovation; it involved navigating one of the most demanding regulatory landscapes in the world. The CE certification under the European Medical Device Regulation (MDR) is not just a label but a legally binding assurance that the device meets stringent standards for safety, efficacy, and quality.
For AI application, the path to CE marking required rigorous validation. The team conducted extensive risk assessments, stress tests, usability evaluations, and cybersecurity audits. A central regulatory focus was the explainability of AI outputs. Clinicians and regulatory authorities alike need to understand how conclusions are reached—especially in high-stakes scenarios. As such, the system’s architecture was designed to ensure every answer could be traced back to its data source and validated reference, ensuring not only clinical utility but also legal robustness.
Achieving CE certification was a turning point in the development journey. It confirmed not just the technical readiness of the product, but its readiness for real-world clinical use. It also positioned the assistant as a pioneering solution in the European medical AI landscape.
The realization of this ambitious project would not have been possible without strong institutional support. The State of Hesse in Germany played a central role in the early stages, offering funding, mentorship, and access to regulatory, academic, and clinical networks. Public innovation programs provided the critical breathing room that allowed the team to iterate, refine, and scale their solution responsibly—without compromising on safety or quality.
The project stands as a testament to the power of collaboration between public institutions and private innovators. Through partnerships with hospitals, academic research centers, and medical societies, the assistant was tested, evaluated, and refined in authentic clinical environments. Physicians from multiple disciplines—including neurology, internal medicine, and pharmacology—contributed to shaping the content, functionality, and performance of the assistant. Their feedback enabled real-world testing and adaptation, ensuring that the system was not only clinically accurate but also practically useful.
Moreover, the involvement of specialists in ethics, data governance, and information security helped establish strong frameworks for data protection, algorithmic fairness, and responsible AI deployment—issues that are becoming increasingly central to the adoption of digital tools in medicine.
Innovation One of the most powerful lessons from this project is that healthcare innovation cannot succeed in isolation. The complexity of clinical care, regulatory oversight, and data security necessitates collaboration across domains. Engineers, clinicians, policymakers, designers, and legal experts must work together, guided by a shared commitment to improving patient care.
This assistant is a product of such a collaborative ecosystem. Its success reflects not only the brilliance of the core development team, but also the diversity that shaped its evolution. It also highlights how regional initiatives—when properly resourced—can lead to nationally and even internationally significant innovations in digital health.
Leadership At the heart of this innovation is Dr. Vera Roedel, an advocate whose leadership has been instrumental in transforming the concept into a functioning, CE-certified product. With a background in both clinical practice and medical education, Dr. Roedel brought a uniquely holistic perspective to the project. She understood not only the technical requirements but also the practical realities of day-to-day medical work.
Her vision centered on user-centric, ethically responsible innovation. She championed the integration of clinicians into the development process and insisted on a transparent, trustworthy system that supports rather than replaces medical professionals. As a female founder in a traditionally male-dominated sector, her leadership also underscores the growing importance of diversity in healthcare innovation—particularly in developing technologies intended for wide and varied use.
Since its founding in early 2023, the project has developed into a scalable digital health platform. Its core product—a CE-certified Clinical Decision Support System (CDSS)—is powered by a domain-specific language model (DSLM) and a Medical Data Analysis Architecture Model (MDAAM). These proprietary systems ensure that only peer-reviewed, curated, and medically validated knowledge informs the AI's output. Unlike general-purpose language models, this one is fine-tuned for the medical domain and constantly updated with new evidence, ensuring its reliability in fast-changing clinical landscapes.
Beyond the assistant itself, the team has launched a CME-accredited educational platform. This offering is designed to help healthcare professionals understand and responsibly adopt AI tools in their practice. It includes training in digital literacy, ethical AI use, and critical evaluation of algorithmic recommendations. In parallel, the team provides consulting services to hospitals and clinics looking to integrate AI into their infrastructure, offering support in interoperability, change management, and clinical pathway adaptation.
A key quality benchmark for medical AI systems is the avoidance of so-called “hallucinations”— that may sound convincing linguistically but are factually incorrect or lack scientific grounding. In the clinical context, where misinformation can have serious or even life-threatening consequences, such inaccuracies are simply unacceptable.
This is why the assistant is built upon an architecture that ensures all content is validated and evidence-based. Every AI-generated response is derived solely from curated medical sources, with clear references and traceable origins. The system is explicitly designed not to speculate or generate assumptions, but to deliver precise, trustworthy information that aligns with current clinical guidelines. This rigorous quality assurance ensures that the assistant operates not creatively, but responsibly and transparently, always in service of patient safety and professional reliability.
The assistant has already gained significant traction, especially among neurologists, and its use is expanding across other disciplines. Strong partnerships with pharmaceutical companies and health institutions are enabling new use cases such as AI-assisted rare disease diagnostics, support for complex treatment planning, and multilingual patient communication tools.
The assistant’s multilingual capabilities also provide major benefits in diverse urban healthcare settings, where language barriers can affect care quality.
Future development will focus on deep integration with electronic health records, enabling personalized, context-aware support based on patient history and lab values. Adaptive learning systems will allow the assistant to learn from user behavior and tailor responses over time, increasing relevance and efficiency. These advancements will remain under ethical oversight and guided by user feedback, preserving the trust the system has earned.
As AI becomes more prevalent in healthcare, ethical responsibility must remain at the forefront. This project exemplifies how responsible innovation is not a constraint, but a foundation for sustainable impact. From data protection and transparency to explainability and equity, the assistant has been built with a clear ethical compass. It prioritizes user autonomy, safeguards patient privacy, and respects the clinician’s role as the final decision-maker.
Rather than replacing the judgment of trained professionals, the assistant enhances it. It acts as a co-pilot—offering guidance, expanding awareness, and lightening cognitive load without ever overriding the human capacity for empathy, nuance, and clinical insight.
The successful development, certification, and deployment of this AI assistant mark a major milestone in the digital transformation of healthcare. It shows that artificial intelligence, when responsibly implemented, can serve as a trusted clinical partner—delivering clarity amid complexity and support when most needed. It also underscores the importance of collaborative ecosystems, visionary leadership, and public investment in bringing meaningful healthcare innovation to life.
As global health systems face rising demand and workforce shortages, AI-powered tools like this will become indispensable. They enable care that is not only more efficient but also more human—helping clinicians focus on the relational and diagnostic aspects of their work while automating cognitive and administrative burdens.
In this future, AI is not an abstract buzzword—it is a practical, ethical, and empowering tool. It’s not about replacing doctors, but empowering them to do their best work. The healthcare professional using AI will be more informed, efficient, and present—bridging the gap between data and empathy, science and care.
References:
https://europa.eu/youreurope/business/product-requirements/labels-markings/ce-marking/index_en.htm
https://www.she-works.de/gruenderinnen-im-portraet/prof-valmed-medizinische-informationen-fuer-aerzte-mittels-ki-sprach-app/2024/07/12/
https://www.nature.com/articles/s41599-024-03811-x
Dr. Vera Roedel, CEO and Co-Founder of Prof. Valmed®, expresses that her passion for medical technology informs the development of trustworthy, CE-certified medical products. She highlights that her core competencies in medical device regulation and strategic communication strengthen the team. Her work reflects her core values: integrity, excellence, and the motivation to make a meaningful contribution through diverse perspectives.
