NVIDIA open models power public sector AI projects across Europe

Why governments turn to open models

According to NVIDIA, open models let governments evaluate them, adapt them to local languages and public services, run them on infrastructure they control, apply their own safeguards and keep sensitive data in controlled environments. Some open models also publish their training datasets, so governments can verify whether the data was licensed, biased or compliant.

Several partners are also building editions of the NVIDIA Nemotron-Personas collection. These datasets sit behind NeMo User Sim, NVIDIA's open source library for population-scale user simulation, which evaluates and trains models and agents in local languages.

European Commission, Poland and the UK

Berlin-based deepset is expanding its Haystack Enterprise Platform, the foundation of AI@EC, the European Commission's shared platform for building and running AI applications on commission-managed infrastructure. The update connects agent development, execution, optimization and governance, and adds tools to compare operating costs, evaluate answer quality and trace problems. NVIDIA Triton Inference Server supports the shared services used to find and rank relevant information.

In Poland, Deloitte is helping build agentic systems for healthcare and employment services. The eHealth Centre is building its Intelligent Services Platform for medical imaging with NVIDIA DGX BasePOD and the MONAI framework. The tool is meant to help physicians analyze X-rays, CT scans and MRI images, detect abnormalities and prioritize urgent cases, reducing waiting times for results.

The Ministry of Family, Labour and Social Policy is developing an agentic factory on NVIDIA accelerated computing infrastructure. Designed to support more than 400 labor offices, it will run models including PLLuM, GPT OSS and Inkling for document search, summarization and digital assistants. NVIDIA is also working with SpeakLeash in Poland and the University of Ljubljana in Slovenia on country-specific additions to its synthetic personas program.

In the UK, London-based Locai Labs released Nemotron-Personas-UK, a dataset of 1 million synthetic personas reflecting the country's demographic and geographic diversity while reducing privacy risk. Potential uses include healthcare evaluation and bias testing across age groups, education levels and rural and urban populations.

Germany, France and Greece

In Germany, Materna is introducing Sinfonie, an on-premises platform combining generative AI, voice interaction and 3D avatars, with NVIDIA Audio2Face available for animation. Built on a project for a German state administration, it lets employees access institutional knowledge through spoken questions while data stays on their organization's infrastructure.

For classified work, secunet, NVIDIA, Red Hat and deepset are developing a sovereign cloud offering built on secunet's SINA Cloud foundation, which includes security components authorized for use with classified information. The planned offering combines NVIDIA accelerated computing and Nemotron open models, Red Hat OpenShift AI, deepset's Haystack agent orchestration, and infrastructure operated by secunet in German data centers.

The Soofi research initiative is developing a family of AI models for German industry and the public sector. That includes Nemotron-Personas-Germany, an open dataset of 1 million records, each with six synthetic personas, grounded in the 2022 census across all 16 federal states and 400 districts. It was created with NVIDIA NeMo Data Designer without real people's private data.

In France, CANUT, the public purchasing body for digital services, selected LINAGORA and Docaposte to provide a sovereign collaborative workspace built on Twake.ai. More than 2,500 eligible organizations, including municipalities, schools and healthcare institutions, can access the framework without running their own tender. NVIDIA technology supports speech transcription, and infrastructure is hosted in France.

In Greece, KIEFER is developing an agentic factory built on NVIDIA DGX SuperPOD with DGX GB200, alongside Sophea open weight models and Greek-language tools for knowledge retrieval, speech and digital assistants.

Codefest and what it means for public bodies

The NVIDIA Open Models Codefest brings together startups, software providers and labs to turn open models into locally controlled solutions for challenges such as medical reasoning, AI tutoring and model governance. The top five teams are Pleias, Zehnlab.ai, Disseqt, Algoritmica GmbH and ABA Technology. Deutsche Telekom and T-Systems provided the compute infrastructure for training and deploying large language models.

For public bodies, the announcements point to a common pattern: models adapted to national languages, run on infrastructure the organization controls, and tested against synthetic populations rather than real personal data. The persona datasets give developers a way to check whether public-service models work consistently across regions, age groups and communities. Several of the offerings, such as the German sovereign cloud, are still described as planned.