I’m a Ph.D. in Artificial Intelligence and Information Theory, with a strong background in Machine Learning and hands-on experience gained through my Ph.D., internships, open-source contributions, and personal projects.
I completed my Ph.D. in December 2024 at Télécom Paris, where I worked on Information Theory and Reinforcement Learning for Mixed Covert and Non-Covert Wireless Networks under the supervision of Prof. Michèle Wigger and Prof. Mireille Sarkiss.
- Reinforcement Learning
- Generative AI (code, text, audio, images)
- AI Agents & AI Security
- AI De-biasing & Fairness
- Applications of AI to Complex Problems
- Strong problem-solving ability with a solid mathematical foundation.
- Experience in deep learning, optimization, and probabilistic modeling.
- Excellent communication, teamwork, and adaptability skills.
In my free time, I actively contribute to open-source projects, especially in Moroccan Darija NLP, with AtlasIA as part of the Core Data Team.
- 🧠 Orthogonalization Neural Networks: A deep learning project exploring vectors orthogonalization in neural networks. It actually (and no so surprising) works. But there are more things to do to force the network to not just do some transformation when the input vectors does not form a basis or are actually already orthogonal (both of these constraints can be added in the objective function).
- 🔓 Crypto Cracker Using Neural Networks: A bit complex to make it work. One must define a good basis so that the neural network can learn. See SALSA: Attacking Lattice Cryptography with Transformers
Let's connect and discuss AI, research, or open-source projects! 🚀