Joseph Barbier

I provide consulting services around: • Automated and reproducible reporting (web, PDF) • Data visualization and storytelling • Analysis pipelines for research and decision-making • Open-source tools for modern data workflows • Accessible PDFs and data visualizations → Get in touch ([email protected]) if you'd like to collaborate!

Context: Dataviz Universe brings together the sites created by Yan Holtz around data visualization. Missions : - Data visualization and information design with Python (mainly Matplotlib) - Creation of open source Python libraries for data visualization (PyPalettes, DrawArrow and PyFonts) - Development of an R package on ecological modelling for the CNRS - Web development (basic HTML/CSS, some React, Quarto) - Other minor tasks

Context: I worked as a freelancer for the Python and R Graph Gallery websites, which are the reference sites for data visualisation in Python and R. They each reference several hundred graphs, explained and reproducible. Missions: - Add new visualizations: matplotlib/seaborn/pandas/plotly and ggplot2/tidyverse/ggplotly - Website enhancement: fixing issues, new packages, IT watch - Share my contributions via Git/Github (top contributor with over a hundred merged pull requests)

Context: Wanteeed is a French startup who provides a browser extension and a mobile app that automatically applies coupon codes and discounts to user's shopping cart, ensuring they get the best price available. Missions: - Data visualization with BigQuery SQL and Looker - Building KPIs - Machine learning with scikit-learn, training via Google Colab - Versioning with Git - PoC of a semantic search engine (embedding, regex, OpenAI API, streamlit)

Tutoring courses in microeconomics for undergraduate economics students (8x1h30)

Context: SHARE is Europe's largest social science panel study, spanning 28 countries for over 20 years. I worked under the supervision of Thomas Salanova, PhD in psychology and data analyst. Mission: To investigate the relationship between socio-economic status, risk behaviours and cancer prevalence. Technical tools: Based on a literature review and numerous available datasets and R, manipulate, clean and exploit data on socio-economic status and cancer. Statistical methods: logistic regression, hypothesis testing, unsupervised learning.