Ouail El khattabi

The Paris Digital Lab is an intensive incubator program where teams of 3–4 students design and deliver three innovative MVPs for partner companies, developed in 7-week Agile sprints to address real-world business needs. 1st project : Subsea Routing Optimization for Offshore Field - TotalEnergies ⚙ Python · Machine Learning · Hexaly Developed a routing solution that ensures connectivity between subsea structures under operational constraints. Leveraged machine learning and the Hexaly solver to minimize total flowline length, improving efficiency and cost-effectiveness. 2nd project : Intelligent Support Chatbot - Cryostar ⚙ Python · LangChain · RAG · OCR · VLM Developing a chatbot for Cryostar employees to answer queries based on internal resources such as user guides and technical documentation, leveraging RAG, OCR, and multimodal AI.

I conducted weekly mathematics tutorials in English for a group of 10 students in the Bachelor of Global Engineering program, a joint initiative between Paris-Saclay University and McGill University.

Tech Projects during my first two years at CentraleSupélec: - LLM-based Coding Assistant for Risk & Reliability Explored optimization strategies for LLMs with RAG and agentic AI, and benchmarked GPT-4o mini vs Hugging Face open-source models; combined approaches improved performance. ⚙ Python · PyTorch · Large Language Models (LLM) - E-Commerce Recommendation System Developed a recommendation engine for fashion products from user-uploaded photos by fine-tuning vision transformer models on a custom dataset. ⚙ Python · PyTorch · Pandas · Computer Vision . Transformers - Classification of Geographical Areas Built a pipeline for preprocessing labeled satellite data, feature engineering, dimensionality reduction, and model selection to categorize new images into six classes. ⚙ Python · Pandas · Torch · Scikit-learn - Tail Assignment – Flight Coverage Optimization Assigned aircraft to flights under continuity, maintenance, and fuel efficiency constraints using linear optimization with heuristics to speed up computations. ⚙ Python · Gurobi - Pac-Man Game (Coding Weeks) Created a Pac-Man game with 3 difficulty levels and random map generation. ⚙ Python · Tkinter · Pygame - Segmentation & Tracking Software Developed automatic image annotation software with YOLOv8 & SAM, fine-tuned on KITTI dataset, integrated with Label Studio for corrections. ⚙ Python · Docker · Computer Vision - Movie Recommendation Web Platform Built a full-stack platform allowing users to rate films and receive personalized recommendations. ⚙ JavaScript · HTML · CSS . SQL · React.js · Express.js

I conducted weekly tutorials for a group of first-year international students under the supervision of Pr. Charles Soussen, covering the courses : Convergence, Integration, Probability and Partial Differential Equations.

During a 5-week internship, I worked as an assembly line operator in an automotive production plant. My responsibilities included installing fender liners on Mokka and DS3 vehicles, both with combustion and electric engines. Specifically, I was in charge of fitting the right-front and right-rear fender liners, ensuring proper alignment, secure fastening, and tightening of bumpers, while maintaining a production rate of 38 vehicles per hour.

CentraleSupélec is one of the top three most prestigious French “Grandes Écoles.” As part of Paris-Saclay University, it contributes to the university’s global ranking, which is 13th worldwide and 1st in Europe according to the Shanghai Ranking. 📚 Relevant Courses: - Machine Learning - Artificial Intelligence (Reinforcement Learning) - Statistics and Learning - Algorithms and Complexity - Measure Theory and Probability - Partial Differential Equations - Advanced Probability - Cloud Computing - Information Theory - Networks and Security - Web Data Intelligence - Economics