mouad el-hachimi

• Master’s thesis on federated learning and automated retraining in edge–cloud with Kubernetes • Built modular FL server and clients with strategy switching via environment variables • Implemented and compared FedAvg, FedProx, and FedRep • Containerized and deployed FL clients as Kubernetes pods at the edge and the FL server as a cloud container • Delivered CNC drilling anomaly detection on real multi-site datasets, with analysis and simulations validating retraining and federation choices

Supporting software quality through code reviews, manual and automated testing, and process documentation. Reviewing pull requests in Bitbucket, creating and executing test cases in Jira/RTM, and automating E2E tests with WinAppDriver, Selenium, and C#. Documenting processes in Confluence and preparing release notes. Also contributed to web app development with TypeScript.

Development of remote solutions in the field of automotive testing. Creation, implementation, and configuration of test data for mobile online services

Abhalten von Sprech- und Fragestunden ---- Betreuung und Unterstützung der Übungsgruppen

Relevant Master's Courses: Database Systems Web Technologies & Cloud Computing Software Engineering for Embedded Systems Industry 4.0 Optimization & Optimal Control Theoretical Electrical Engineering Numerical Methods Multidimensional Functions in Electrical Engineering

Bachelor Thesis: Evaluation of Reinforcement Learning Frameworks for Industrial Applications Analysis and evaluation of RL frameworks such as OpenAI Gym, TensorFlow, and Stable Baselines3. Implementation and comparison of algorithms such as Q-Learning, DQN, Monte Carlo, and PPO in Python. Execution of experiments in simulated environments (e.g., CartPole, Blackjack) for performance assessment. Fine tuning and model stability analysis using TensorBoard.