Source proof for Cliniques universitaires Saint-Luc deploys an on-premises Qwen3-235B assistant inside its Epic EHR, adopted by 1,028 users

…eval, and note drafting, which together accounted for over 70% of interactions. Following the pilot, the system was deployed hospital-wide and adopted by 1,028 users who generated 14,910 conversations over five months, with more than half of clinicians using it at least weekly. Usage remained concentrated on information access and documentation support, in…

…his burden, yet real-world deployments in live hospital systems remain limited. We implemented a secure, GDPR-compliant, on-premises LLM assistant integrated into the Epic EHR at a European university hospital. The system uses Qwen3-235B with Retrieval Augmented Generation to deliver conte…

…s LLM assistant integrated into the Epic EHR at a European university hospital. The system uses Qwen3-235B with Retrieval Augmented Generation to deliver context-aware answers drawing on structured patient data, internal and regional clinical documents, and medical literature. A one-month pilot with 28 physicians across nine specialties demonstrated high…

…patient data, internal and regional clinical documents, and medical literature. A one-month pilot with 28 physicians across nine specialties demonstrated high engagement, with 64% of participants using the assistant daily and generating 482 multi-turn conversations. The most common tasks were summarization, information retrieval, and note draft…

…ese tools useful and safe in day-to-day workflow. To support our GenAI efforts, we invested in a server with 8 NVIDIA H200 GPUs, allowing us to host the largest available models, such as DeepSeek V3. In addition to our inference capabilities, this server provides the necessary c…

…ft, Writing – review & editing * E-mail: maxime.griot@uclouvain.be Affiliations Cliniques Universitaires Saint-Luc, Woluwe-Saint-Lambert, Brussels, Belgium, Université Catholique de Louvain, Louvain Research Institute in Management and…

…ft, Writing – review & editing * E-mail: maxime.griot@uclouvain.be Affiliations Cliniques Universitaires Saint-Luc, Woluwe-Saint-Lambert, Brussels, Belgium, Université Catholique de Louvain, Louvain Research Institute in Management and…

…ft, Writing – review & editing * E-mail: maxime.griot@uclouvain.be Affiliations Cliniques Universitaires Saint-Luc, Woluwe-Saint-Lambert, Brussels, Belgium, Université Catholique de Louvain, Louvain Research Institute in Management and…

…his burden, yet real-world deployments in live hospital systems remain limited. We implemented a secure, GDPR-compliant, on-premises LLM assistant integrated into the Epic EHR at a European university hospital. The system uses Qwen3-235B with Retrieval Augmented Generation to deliver conte…

…uthors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract Electronic Health Records (EHRs) have improved access to patient information but substantially increased clinicians’ documentation workload. Large Language Models (LLMs) offer a potential means to reduce this burden, yet…

…s LLM assistant integrated into the Epic EHR at a European university hospital. The system uses Qwen3-235B with Retrieval Augmented Generation to deliver context-aware answers drawing on structured patient data, internal and regional clinical documents, and medical literature. A one-month pilot with 28 physicians across nine specialties demonstrated high…

…ble and affordable implementations without compromising data governance [ 33 ]. Our software stack is depicted in Fig 1 based around vLLM [ 34 ], which provides the LLM inference capabilities using the same protocol as OpenAI, allowing us to use existing libraries and me…

…eval, and note drafting, which together accounted for over 70% of interactions. Following the pilot, the system was deployed hospital-wide and adopted by 1,028 users who generated 14,910 conversations over five months, with more than half of clinicians using it at least weekly. Usage remained concentrated on information access and documentation support, in…

…chnical implementation, measure user adoption, and characterize usage patterns. We deliberately did not measure clinical outcomes or time savings, as this would require a controlled study design inappropriate for an initial deployment evaluation. Infrastructure. LLMs require extensive computing capabilities, especially to su…

…s LLM assistant integrated into the Epic EHR at a European university hospital. The system uses Qwen3-235B with Retrieval Augmented Generation to deliver context-aware answers drawing on structured patient data, internal and regional clinical documents, and medical literature. A one-month pilot with 28 physicians across nine specialties demonstrated high…

…s critical for downstream systems that rely on predictable, well-formed inputs. We use LiteLLM to manage access control and system reliability by enforcing usage limits and distributing requests across multiple vLLM instan…

…nature of the documents allows for more accurate searches, especially semantic. We therefore opted for full-text semantic search using Apache Solr instead of vector databases. The documents are chunked based on level 1 and 2 titles present in the document…

…minimalist chat-based interface accompanied by a list of previous interactions. We built on the open-source OpenWebUI interface, simplifying it for clinical users while preserving its core chat functionality [ 39 ]. Originally designed for advanced users seeking fine-grained control ove…

…ese tools useful and safe in day-to-day workflow. To support our GenAI efforts, we invested in a server with 8 NVIDIA H200 GPUs, allowing us to host the largest available models, such as DeepSeek V3. In addition to our inference capabilities, this server provides the necessary c…

…apabilities, especially to support thousands of users in a university hospital. The GenAI strategy of our institution is to host medical applications on-site and not to rely on vendor services that do not allow for the level of customization and validation required to mak…

…controlled environment and filters it to retain only curated, relevant subsets. The core inference and retrieval pipeline operates entirely within the hospital’s internal network and does not require outbound internet connectivity. Access to Microsoft SharePoint documents hosted on a private Azure tenant is me…

…shed to all users or rejected. Compliance with legal and regulatory frameworks. The system was notified to the Belgian Federal Agency for Medicines and Health Products as an in-house medical device under Article 5(5) of the European Medical Device Regulation (MDR) [ 41 ]. This provision permits healthcare institutions to develop and use custom medica…

…several obligations for high-risk AI systems being phased in over time [ 42 ]. Given the clinical context of use, we proactively aligned the system’s development and governance with the requirements applicable to high-risk medical AI systems. In particular, we implemented: Clear documentation of intended use, model behav…

…m corresponds to a class IIa medical device under the MDR classification rules. With respect to data protection, the system was designed in full compliance with the General Data Protection Regulation (GDPR) [ 40 ]. No personal data leave the hospital’s infrastructure, and access to patient inf…

…eval, and note drafting, which together accounted for over 70% of interactions. Following the pilot, the system was deployed hospital-wide and adopted by 1,028 users who generated 14,910 conversations over five months, with more than half of clinicians using it at least weekly. Usage remained concentrated on information access and documentation support, in…

…s over five months, with more than half of clinicians using it at least weekly. Usage remained concentrated on information access and documentation support, indicating stable incorporation into everyday clinical workflows. Feedback volume decreased compared with the pilot, suggesting that routine use…