Developed this tool for the Macron Presidential Campaign. The tool did a weekly scrape of the principal regional outlets for every region in France as well as the leading national newspapers (used as benchmark). The results were shared in the internal communication note sent to the highest-ranking members of the campaign.
As part of my Data Science Fellowship at the U.S. Bureau of Labor Statistics, I developed a tool that automated days of analysis and visualization to quickly extract important highlights and trends from new or revised data. The tool is hosted on BLS' server and isn't available to the public.
(These GIFs were taken with authorization from superiors)
The 1st feature I developed was the Revisions Feature
The 2nd feature I developed was the Outliers Feature
In this project, I applied Machine Learning models to my Goodreads Data to identify my next favorite books 📚among a list of more than 10 000. The project's code can be found here 👨💻
Delphes is an Open-Source Deep-Learning website trained on more than 150 000 tweets from Members of the European Parliament (MEP) that matches European citizens with their closest European Political Group and MEP, based on a simple text-input or Twitter username.
Check out my Presentation of Delphes at Le Wagon's Demo Day
(The video is already set on my presentation to save you time)
TwittLists analyzes millions of public user-generated Twitter lists to create a ranking of the best Twitter accounts on a given topic ranging from Chess to French Mayors.
In November alone the account reached 500K people!
