Data & AI consultant and trainer · Founder of Data & AI Organization
10 years turning data into decisions, and teaching others to do the same.
🚢 Live: Portoria.app
The US Harmonized Tariff Schedule, made readable. Every one of its 29,861 codes has its own page with an AI-written overview, its duty rates and the notes of its chapter.
- Always up to date. A daily pipeline checks the US International Trade Commission for a new revision, then updates the data, writes overviews for new or changed codes, rebuilds the site and deploys it, with no manual step.
- Change alerts. Follow any code and get an e-mail when a new revision changes it.
- Built with Python (static site generator), Gemini, Cloudflare Workers and D1, GitHub Actions.
Through Data & AI Organization, I help small and mid-sized European businesses turn their data into automated, reliable processes. Fixed-price projects, delivered in English or French.
- Data collection & preparation: web scraping, data extraction and pipelines that gather prices, listings, leads or business data automatically, then clean and structure it.
- Dashboards, analytics & machine learning: always up-to-date dashboards that replace manual reports, plus forecasting and predictive models.
- AI automation, AI agents & chatbots: AI that processes e-mails, invoices and documents, agents that handle multi-step tasks, and chatbots that answer from your own documents.
- AI consulting & training: an audit of where AI would really pay off, GDPR and EU AI Act readiness, and hands-on training for your teams.
I design and teach data, AI and programming courses, in higher education from first year to Master's level, and in professional training, from managers to executives. More than 500 apprentices trained in computer science, data and AI.
- formation-r: a hands-on introduction to R for beginners, from installing the workstation to statistics, with an exercise and its solution for every session, on real French education data.
- Subjects I teach: AI development and AI with Python · data science for full-stack developers · scraping and data cleaning · data visualisation · programming paradigms · application and system architecture · maths and statistics for computing · web, CMS and no-code tools.
| Project | What it does | Stack |
|---|---|---|
| France_Competence_db | Relational database of French professional certifications (RNCP and RS), linked to job codes and to the organisations that certify and teach them | Python, SQLAlchemy, SQLite, scraping |
| classement-prefecture | Ranks French prefectures by their Google Maps ratings, with an interactive map including overseas departments · live app | Streamlit, Altair, GeoPandas |
| morocco_admin_geography | Ready-to-use dataset of Morocco's regions, prefectures and communes in French and Arabic, with the scraper that builds it | Python, CSV, JSON, Excel |
| a-repondre | Collects the questions people ask Google about a topic, to plan content, with a cap on API costs | Flask, SerpAPI |
| malaria-detection-app | Classifies blood-cell images as infected or not with a CNN, packaged in Docker with a deployment pipeline to Azure | TensorFlow, Flask, Docker, GitHub Actions |
- Data & AI: Python, R, SQL, machine learning, LLMs and RAG, web scraping
- Apps & cloud: Flask, FastAPI, Streamlit, Docker, Cloudflare Workers, Google Cloud, Azure, GitHub Actions
- Digital analytics & marketing: Google Analytics, Google Tag Manager, Google Ads, SEO and SEA
- eBusiness Consultant, Business & Decision: measurement plans for e-commerce sites, restructuring of Google Analytics and Tag Manager 360 accounts, international coordination.
- Digital Analytics & Marketing Consultant, Agence SW: Google Ads and Facebook Ads campaigns, web performance measurement.
- Regional winner, Grand Est, PEPITE 2020 entrepreneurship competition (France)
- 1st prize, Coup de Pouce 2020 by Yzico & PeeL (Nancy)