OakLabs is a mobile encyclopedia application on the Pokémon universe: a Pokédex. The main functionality is the recognition and classification of Pokémons via the camera thanks to artificial intelligence. There is also the list of all Pokémons and their characteristics, management of badges for players, and options: choice of language and theme.
This important project, realized in group of 5 in agile scrum method, is part of my last year training at ENSSAT, in the software engineering module. On my side, my tasks were the graphic charter of the application, all the user paths and the mock-ups (light and dark themes), as well as the front-end development of some pages (for example the Pokémon detection modal, the Pokédex list and the page of a Pokémon).
This project is part of the Orange Phone mobile anti-spam application, and more specifically the reverse directory which allows companies to display their name and the reason of their call for Orange Phone users. The goal of this end-of-studies project at Orange is to implement a dual authentication system for businesses who want to call their customers, in order to prevent their numbers from being usurped.
Orange Antispoofing is realized by another apprentice and myself, in agile scrum method, from A to Z. The 2 main steps are the design phase and the study of the competition, then the development phase. To do this, we had to rework the existing parts of Orange Phone existing parts: the web portal (in Flutter), the Android application, the iOS application, and finally add the whole back-end API part with a database in MongoDB and a link with Firebase.
As part of my engineering degree, it was necessary to do a 3-month internship abroad. Done at the University of Cagliari in Sardinia, the project was a research topic . The aim was to determine whether developers preferred SQL or NoSQL languages. To do this, I based my research on the developers' comments on the GitHub repositories of 10 major database management systems.
The first part was to get these comments in Python, thanks to the API proposed by GitHub. After retrieving about 50 000 of them, I performed several sentiment analysis with the with the Transformers artificial intelligence library as well as other analyses (evolution of the number of comments per year, sentiment for each DBMS...). Finally, I developed a small web site to list these researches. I also wrote a research paper on this project.
