Clé brings language and image models to patient care, research, teaching and administration – running entirely on Charité-owned servers.
Charité Switches On Its Own AI Platform, Built by a Berlin Spin-Off
NewsShare on:
Charité – Universitätsmedizin Berlin has launched its own AI platform. "Clé" is available to staff in patient care, research, teaching and administration, and runs entirely on Charité's own servers. According to the hospital, that setup is so far unique among Germany's university hospitals.
"To our knowledge, Charité is the first university medical centre in Germany to run its own AI platform with several language and image models entirely on its own infrastructure – not as a pilot in individual departments, but across the board," says Prof. Alexander Meyer, head of the Institute for Artificial Intelligence in Medicine (IKIM) and the project's initiator.
One environment, four areas of use
Clé combines several interchangeable AI models, a chat interface, processing of text, images and PDF documents, and dedicated knowledge collections. Via APIs, researchers and departments can embed the models into their own applications and workflows. Central login, technical safeguards and usage logging are designed to keep security standards consistent.
The key difference to external services: staff can work with internal documents and business information without transmitting any of it to providers outside Charité. The rollout is accompanied by mandatory training and phased access.
"A thoroughly secure AI infrastructure is a decisive step forward for all of Charité, particularly given the current debates," says Prof. Heyo K. Kroemer, Chairman of the Board at Charité. "It allows us to push digital transformation and lay the groundwork for AI-supported processes."
Relieving the workload – explicitly not diagnostics
In patient care, Clé is meant to support clinical-administrative tasks such as structuring documents and preparing information. The goal is to cut routine effort and free up time for patients.
The platform is explicitly not a medical device. "In its currently approved form, Clé is not intended for diagnoses or treatment decisions," Meyer stresses. "Clinical responsibility stays with the treating professionals." AI outputs, he adds, always need to be reviewed critically.
A pilot project shows how much leverage sits on the administrative side: converting documents into structured data dropped from an average of 30 minutes to two. Across 1,800 documents, that adds up to 105 working days saved. Other use cases include knowledge bases, automated translation, form processing, and revenue and HR management.
In research, the APIs allow teams to build their own applications and analyse large volumes of text and documents securely. The platform also becomes a research object in its own right, making it possible to study under controlled conditions how language models can be deployed safely and effectively in medicine. In teaching, it is being used to prepare course materials, among other things.
Next stage: AI agents
The platform was designed from the outset as a foundation for agent systems that handle multi-step tasks independently – evaluating documents, merging data, preparing drafts, automating clinical processes. The first such applications are already being built via the API in research and IT.
"In the next expansion stages, agents should work in a secured environment directly in the interface and be connected to Charité systems step by step," Meyer says. "The same principles apply as today: data and control stay within Charité, and responsibility stays with people."
Berlin ecosystem: built by Ascléa
Clé was developed by Ascléa GmbH, a spin-off of Charité and the German Heart Center of Charité. The platform is operated together with IKIM and Charité's IT division on in-house infrastructure, with the High Performance Computing team maintaining the supercomputer the models run on. The Center of Technical Excellence supports the build-out; the Governance, Compliance and Data Protection division handles the regulatory side.
Ascléa is a familiar name in Berlin's ecosystem. The company emerged from x-cardiac, known for AI tools that predict complications after heart surgery, and closed a Series A round in July 2026 with new backers Quercus and Catalpa Ventures alongside existing investors IBB Ventures and a group of business angels. Two CE-certified AI medical products are already on the market.
Transparency note from Charité: Alexander Meyer is a co-founder, shareholder and one of the managing directors of Ascléa. Charité holds a stake in Ascléa through its technology transfer partner Ascenion, uses the platform free of charge and shares in the company's revenue; Charité makes no payments to Ascléa. The contractual arrangements were negotiated and approved by the responsible Charité bodies.
Source: press release by Charité – Universitätsmedizin Berlin, 17 September 2026.