Data Engineering/Science Internship for Public Health Surveillance (EUROMOMO)
Référence du poste : DATA-STA-2026-10
Cette offre vous intéresse ?
Nous rejoindre
Vous avez envie de faire carrière au sein d'un organisme public qui a pour mission de protéger efficacement la santé des populations ? Rejoignez-nous.
Santé publique France is France’s national public health agency. A public institution under the authority of the Minister of Health, formed through the merger of several public institutions and established by Ordinance 2016-246 of April 15, 2016, the agency works to promote public health. As a scientific, expert, and public health safety agency, its missions include:
- Epidemiological observation and surveillance of the health status of the population;
- Monitoring health risks threatening the population;
- Promoting health and reducing health risks;
- Developing prevention and health education;
- Preparation for and response to health threats, alerts, and crises;
- Issuing public health alerts.
The agency is organized into 12 departments—some scientific, some cross-functional, and some providing operational support.
The agency’s strategic priorities and work program, established by its Board of Directors, are organized into three areas: Strengthening the capacity for anticipation and rapid response to health threats; Measuring and assessing the extent of diseases and risk factors to guide their prevention and control; Strengthening the health impact of all public policies, as well as health prevention and promotion.
Data Support, Processing, and Analysis Division (DATA)
Internship Topic
Santé publique France monitors all-cause mortality on a weekly basis using vital statistics data provided by INSEE, relying on the European EuroMOMO model to estimate the expected number of deaths and detect any excess mortality. An R Shiny application developed in 2015 allows these analyses to be visualized on a daily and weekly basis, for all ages and by age group, at the national, regional, and departmental levels. It is based on a sample of approximately 5,000 municipalities (accounting for 84% of national mortality). In the first half of 2027, an expansion of the sample to approximately 12,000 municipalities will enable the use of another major analytical dimension: the type of place of death (home, hospital, nursing home). This application is updated daily with new data received from INSEE.
The internship aims to modernize the R-Shiny application and enhance the analyses by: integration of new strata (gender, type of place of death), integration of results from a nowcasting method to account for data transmission delays, mapping by type of place of death, and excess mortality indicators that are more refined than simply exceeding a threshold. It also includes a redesign of the code architecture to ensure that all Santé publique France applications are standardized.
The internship is part of a broader effort by the DATA department to automate the creation of Shiny applications using agent-based AI: the development and modernization of the application will rely on autonomous code agents (such as Cline, connected to Mistral AI models), which are set to become a standard development method for the agency’s applications. The internship will therefore involve creating and improving “skills”: libraries of reusable guidelines, contexts, and procedures that guide these AI agents in generating, refactoring, and documenting R/Shiny code. This approach aims to accelerate development, ensure the reliability of outputs, and capitalize on the unit’s methodological expertise by creating assets that can be reused by other teams within the agency.
Internship Responsibilities
Getting Started and Assessing the Current Situation
- Understand the mortality surveillance system (INSEE data, reporting deadlines, EuroMOMO principles) and the model outputs made available.
- Audit the existing R Shiny application (code, dependencies, performance, usability) and propose a roadmap for improvement.
- Familiarize yourself with the AI-assisted development environment (Cline, Mistral AI models) and its usage framework (security and confidentiality).
Redesign of the application architecture
- Adapt the application to the new data volumes and format, and optimize its performance.
- Structure and document the code (reproducible project, tests, version control), aligning it with other Santé publique France applications.
AI-assisted development and creation of “skills”
- Use the code agent (Cline / Mistral) to accelerate refactoring, module generation, testing, and documentation, subject to systematic human review.
- Design, document, and ensure the reliability of “skills” (instructions, business contexts, code patterns, validation procedures) for recurring project tasks.
- Contribute to a reusable development framework for the agency’s future applications by measuring the impact of skills on development quality and speed and by sharing best practices with other teams in the DATA division.
Integration of New Stratifications
- Add results stratified by gender and type of place of death at the geographic levels defined with the team, using appropriate visualizations (graphs, tables, comparisons between strata).
Mapping
- Develop interactive maps of mortality and excess mortality by type of place of death and according to other strata.
Excess Mortality Intensity Indicators
- Represent levels of excess mortality intensity—beyond simply exceeding a threshold (e.g., by intensity classes)—in collaboration with the team’s statisticians.
Validation, Documentation, and Rollout
- Test the application with users (surveillance team, regional units) and incorporate their feedback.
- Draft technical documentation and a user guide, then present the results at debriefing meetings.
Write a report on lessons learned regarding the use of generative AI (benefits, limitations, and recommendations for the agency).
Our latest news
news
Call for Nominations to Fill Vacancies on the National Reference Centers...
news
Public Health France Conference: November 9, 2026
news