Back-End Developer & Data Engineer

During a six-month internship at Hydratis, a hydraulic research laboratory, I had the opportunity to work on the early stages of an innovative solution aimed at detecting floods before they occur, with the goal of evacuating the population before they are affected. I built an Elasticsearch database to store meteorological, hydraulic, and soil data collected from various sensors at different locations. Additionally, I developed a Python server to retrieve and send this data to Elasticsearch. I also created real-time data visualization dashboards using Kibana and designed a graphical user interface with Tkinter.

Teaching Assistant

For one year, on a schedule of two days per week, I assisted the pedagogical team at Epitech, supporting first and second-year students. I provided guidance during the intensive C and C++ coding pools, answering their questions, organizing workshops, and evaluating student's work.

R&D Developer

During a four-month internship at Free Pro, I initially worked on the development of a 5G SA private demonstration. For this project, I first created a private network and connected all my devices (Raspberry Pi, Niryo Ned 2 robot arm, iPad, computing machine, camera) in different ways (WiFi, manually configured SIM cards, cable). I developed a Python server to control the demo and the robot, fine-tuned a YOLO v8 model for image classification, and built a React Native application to control the entire demonstration. Finally, I worked on other side projects, such as getting a Pepper robot running and creating an internal mini ChatGPT.

AI Developer

During a two-year work-study program at Free Pro, I worked on the creation of a knowledge base exploitation platform based on RAG. I developed a NestJS server using HTTP/WebSocket and based on a microservice architecture, managing all communications within the platform. I contributed to research on RAG methods using Transformers, LangChain, Llama Index, Chroma DB, and Qdrant. I developed various microservices in Python to build the RAG and one for creating LLM agents based on Smolagents.