Software Engineer at Arista Networks, I specialize in data-science and machine-learning. Though I am open to any new technology, the languages I am most familiar with are C++, Python, JAVA, and C. My interests are various and involve -- among others -- scientific and technological innovation, music, literature, and history of Art.
Data analysis projects involving Machine Learning techniques using the Python scientific ecosystem, Keras, and the Go language.
Organising practical C++, JAVA, C, SQL, and UNIX courses and projects for 400+ third-year students as part of a team of 30 assistants. Our job was to prepare the subjects, organize courses, help the students understand programming and debug their code, and check their work. I was responsible of the Chess project (complete Chess game engine with an AI in C++ 14), and of the Formula One project, a race car driving AI. I was also the quality manager, in charge of checking all the documents (courses, subjects, exams...) provided to the students to ensure that the quality (grammar, content, difficulty) was sufficient.
- Development of a multimedia application controlled with the hands using a Leap Motion sensor.
- Development of an IFC 3D model import plugin for the Minko 3D Engine (C++).
- Development of an interactive virtual visit app using Oculus Rift DK2 base on the Minko 3D Engine (C++).
- Implementation of clipping planes inside the Minko 3D Application to allow users to see inside 3D models (C++).
Teaching UNIX and C Programming to a thirty students classes. My mission as Teaching assistant at EPITA was to supervise second-year practice class. I taught UNIX, GNU/Linux and C Programming to a class of thirty students. This gave me the opportunity to improve my teaching skills my technical skills nd my pedagogical abilities.
The SYDER is a town union responsible for energy and public lighting of a great part of the department of Rhône (France). My mission inside the company was to develop a Web App allowing town councils to consult their charges, and to simulate their budget knowing the ongoing projects. I also improved the Web Design of the Extranet portal and I developed a small cartography system based on OpenLayers and OpenStreetMap.
- harmonic, rhythmic, and melodic analysis of MIDI files
- machine-learning to learn new music genres using the analysed data
- generation to compose music using the trained data
- A compiler, which reads a text dictionary, builds a Trie, and serializes it into a binary contiguous file that allows direct search without reconstructing the Trie in RAM.
- A spell-checker that directly reads the dictionary on SSD using mmap to allow a great number of requests per second without spending time reconstructing the trie in RAM.