Philippe M.

I designed and validated the ETL pipeline that aggregates and reconciles digital inclusion data from multiple sources (Coop, Aidants Connect, National Mapping, SIRENE). My work combines large-scale data analysis — consistency audits on hundreds of thousands of records, duplicate and anomaly detection, correction of confidentiality breaches — and automated enrichment techniques: fuzzy matching, deduplication, geocoding, and batch API enrichment (BAN/IGN, SIRENE/INSEE). I also led a data model redesign that improved the accuracy of indicators exposed via the public API, with a constant focus on analytical rigor and data reliability.

I contribute to the development of Mon Inclusion Numérique, a public platform that supports digital inclusion actors across France, within the Transverse Sept team at ANCT.

Fullstack development of the Mon Inclusion Numérique platform (Next.js, PostgreSQL).

Building the data warehouse that aggregates and unifies heterogeneous sources from the digital inclusion ecosystem — the foundation for monitoring, reporting, and decision-making.

Implementing engineering best practices, particularly in CI/CD, to ensure reliable deployments and manage technical debt.

An environment with high public impact, short cycles, and daily practice of AI-augmented development with Claude Code.

I contributed to the development of Yaso's technical infrastructure and data processing workflows, with a strong focus on security — a structuring constraint in the context of health data.

Designed the infrastructure around Keycloak (authentication) integrated with Parse Server (BaaS), to ensure secure access aligned with business and regulatory requirements.

Extended Parse Server via Node.js modules to adapt the authentication service to the specific needs of the health sector.

Implemented secure data flows and optimized access controls to protect sensitive patient data.

Developed the user interface in Elm, in pair programming with the Product Owner, to maintain close alignment between delivered features and business needs.

Developed data processing workflows in Python, compliant with regulatory requirements.

Automated deployment for onboarding new hospitals, adhering to a high security standard.

Built CI/CD pipelines under GitLab-CI, integrating Docker, Process Compose, Devbox, and git-hooks, with automated tests and deployments.