Rémi Coudert

Komugi is a french B2B startup developing an APS (Advanced Planning and Scheduling) SaaS solution for industries. We provide advanced scheduling with scenarization, load/capacity computations, machine and operator management, turnover computation etc... My roles at Komugi are Fullstack and DevOps engineer. We work in Agile with the SCRUM methodology and my User Stories can involve for the Backend part working in depth on the scheduling algorithm, on database access, optimizations, API development, Data importation through files etc... On the frontend part, I usually update or create graphs, update data from the backend, or add new logic As a DevOps engineer I also make use of Terraform to create and maintain our client environments on AWS, update logs, or develop new capabilities. For example, I implemented an event based automatic deployment of a docker computation service in order to use more powerful ECS pods for cheaper. Previous implementation ran on the 24/7 API server.

Following my Master thesis, held a Research engineer position at the Information Security and Privacy Lab, Lausanne University, lead by Prof. Kévin Huguenin. As part of the lab team, my work consisted in helping with various internal papers (writing, data visualization etc...) and projects Most of my work was done using Python and libraries such as Pandas, Matplotlib, OpenWPM or scikit-learn

During my Master Thesis titled "Automatically detect Dark patterns in cookie banners" at PRIVATICS lab (INRIA Grenoble), I developed a comprehensive tool to detect and analyze dark patterns in cookie consent banners across the web. The research involved crawling over 100,000 websites from the Tranco top list using the OpenWPM tool. I extracted DOM HTML components and developed a scoring system to assign likelihood scores for identifying the outermost components containing cookie banners. From this large-scale analysis, I successfully identified 9,344 cookie banners and conducted detailed analysis of dark patterns - UI design patterns that exploit human cognitive biases to manipulate users into accepting cookies against their best interests. The research findings revealed that 96% of the analyzed cookie banners contained at least one form of dark pattern, highlighting the widespread nature of manipulative design in web privacy interfaces. The complete methodology and detailed analysis are available in my detailed analysis and thesis report.

I work(ed) as a (student) Teaching Assistant for several EPFL courses such as : - AICC (Advanced Information, Computation and Communication), 1st year students - Programmation Orienté Système (System Oriented Programming - C), 2nd year students - Projet Programmation Orienté Système (Project in System Oriented Programming - C), 2nd year students - Computer language processing (Compilers bases), 3rd year students - Introduction to programming , 1st year students

6 month long engineering internship at Logitech. I worked on a project involving Big Data analytics, Machine Learning and E-Sports in the R&D department of Logitech Lausanne (HQ). The goal was to find the future pro players using game data.

I participated to Google Summer of Code 2017 with the Scala Native organization. Scala Native is an ahead of time compiler for Scala, targeting LLVM. My work was to port java Threading to Scala porting existing Java code and mixing it with C code. I had, for instance, to build Java Threads on top of C pthreads.