Antoine Dersy

Supélec engineer, data scientist at WhatsApp (Meta), Criteo, and Google, I work with you to leverage your data.

Passionate about business as the end goal and technology as the means, I am convinced that there are simple, effective, and inexpensive methods to help companies gain competitiveness through data.

I offer a complete or partial data valorization service in three sequential phases: Description-Analysis-Prediction, with a cross-cutting focus on automation.

First, we will evaluate your existing datasets together. Then, with the business teams, I will propose data-driven methods to meet their specific needs. Finally, once the data is analyzed and the link with the business is clarified (two necessary conditions for success), we can develop a machine learning POC and consider its industrialization.

1/ Description and Visualization

- Exploratory analysis, descriptive statistics [R: dplyr + ggplot or Python: Pandas + Seaborn/Plotnine/Bokeh]

- Database quality and consolidation

- Data visualization [Tableau, R Shiny]

2/ Business Analysis (Business Intelligence)

- Data mining and database cross-referencing to gain insights

- Dashboard development

- Experimental analysis protocols (analysis bias, A/B test design)

- Statistical tests (confidence intervals, p-values)

- Explanation and interpretation of complex business phenomena, particularly using visualizations

- Coordination and presentation of results to technical and non-technical teams

3/ Prediction

- Optimization of BI methods from phase 2

- Classification (clustering, segmentation, anomaly detection, fraud) [Scikit, XGBoost, Stan, Keras, TensorFlow]

- Prediction (scoring, forecasting, churn)

- Natural Language Processing (NLP) [NLTK]

X/ Automation and Maintenance

- Job/ETL creation [AWS, HDFS, Spark]

- Code maintenance [Git]

- Web Scraping [Selenium, bs4]