ThinkDeep's DeepBrain AI agents help automate public services for the French government
ThinkDeep AI developed DeepBrain, a multi-agent assistant platform for France's Ministry of Economy and Finance and Ministry of Defense, built on NVIDIA AI Enterprise (NIM, NeMo Retriever, NeMo Guardrails, Llama Nemotron models) running on-premises on NVIDIA DGX H100 systems to keep sensitive data within government infrastructure and comply with the EU AI Act. DeepBrain agents process millions of PDF documents, scanned images, schemas and videos for use cases including fraud detection and legal document processing. Document retrieval time fell from two days to two minutes, a typical 100-user deployment saves 1,000 work hours per month, and the Ministry of Finance has saved €2 million deploying DeepBrain at scale to 10,000 employees.
Overview
ThinkDeep AI developed DeepBrain, a multi-agent assistant platform for France's Ministry of Economy and Finance and Ministry of Defense, built on NVIDIA AI Enterprise (NIM, NeMo Retriever, NeMo Guardrails, Llama Nemotron models) running on-premises on NVIDIA DGX H100 systems to keep sensitive data within government infrastructure and comply with the EU AI Act. DeepBrain agents process millions of PDF documents, scanned images, schemas and videos for use cases including fraud detection and legal document processing. Document retrieval time fell from two days to two minutes, a typical 100-user deployment saves 1,000 work hours per month, and the Ministry of Finance has saved €2 million deploying DeepBrain at scale to 10,000 employees.
This entry has 13 published fields tied to exact passages in an immutable source capture.
Inspect the highlighted sourceThe challenge
The French administration handles over 100 million documents annually to provide tailored responses to citizen inquiries, requiring public servants to stay abreast of changing regulations, maintain accuracy, and cite specific legal references — all while processing huge amounts of legal documentation in real time and ensuring data confidentiality, security, and compliance with the European AI Act.
The solution
ThinkDeep AI developed DeepBrain, a multi-agent assistant platform that lets public servants build personal AI agents connected to a wide range of data sources, cross-referencing information to find patterns for use cases like fraud detection and legal document processing. DeepBrain runs on the NVIDIA AI platform on premises — including NVIDIA H100 and H200 Tensor Core GPUs, NVIDIA DGX systems, and NVIDIA AI Enterprise software such as NIM microservices, NeMo Retriever, Llama Nemotron models and NeMo Guardrails — keeping data on the customer's own infrastructure at all times for security and confidentiality. ThinkDeep used NVIDIA NIM to deploy LLMs on premise, NVIDIA Dynamo to optimize inference speed, and the NeMo framework to integrate LLMs and vision language models.
Reported business value
DeepBrain reduced document retrieval time from two days to two minutes. Implementation took just one week to set up a DeepBrain instance and configure connectors. A typical 100-user deployment saves 1,000 work hours per month, and the Ministry of Finance has saved €2 million by deploying DeepBrain at scale to 10,000 employees.
Sources
Open any source and check the claim yourself — that is the point of the register.
This record was researched and written with AI assistance, and its claims were checked against the sources above. (EU AI Act art. 50 transparency notice.)
Other government & public sector entries in the register.
GovTech unlocks data insights to improve nationwide services with Databricks
Singapore's Government Technology Agency (GovTech) migrated from an on-premises dashboarding system to the Databricks Data + AI Platform on AWS with Unity Catalog and Delta Lake, cutting dashboard creation time from 90 to 30 days, democratizing data across 50% of corporate divisions in the first year, and saving 8,000 labor hours annually.
Estonia rolls out Bürokratt, an AI-guided virtual assistant network for public services
Bürokratt is a network of chatbots deployed on Estonian public sector institutions' websites, letting people obtain information from institutions and use public and information services via virtual assistants. It is a state-created, AI-based digital assistant that helps institutions deliver modern, efficient, around-the-clock customer service using large language models.
VA Advances Healthcare Insights With AI
The U.S. Department of Veterans Affairs leverages Databricks Data Intelligence to modernize healthcare analytics for millions of veterans, unifying massive distributed datasets into a single secure environment and streaming petabytes of health data in real time, reducing processes that once took hours to seconds. Databricks enables AI and large language models to detect risk early and enhance governance.
Austrian Academy of Sciences unlocks Ancient Greek with Mistral
The Austrian Academy of Sciences (OeAW), together with its Austrian Archaeological Institute, partnered with Mistral and services partner Reply to build Apollo, described as the first advanced large language model for Ancient Greek. Apollo is trained on a specialized corpus of 600 million words of historical Greek text plus tens of thousands of published inscriptions and papyri, helping researchers reconstruct damaged texts and identify thematic connections across collections. The OeAW reports Apollo turns work that once took years into hours, addressing over one million unread Greek papyri worldwide, with future phases planned for semantic search and handwritten inscription decipherment.
Was this helpful?
Your feedback helps us improve our use case database

