Stage ingénieur : IA Adaptative de confiance : Apprentissage Fédéré et Continu sous Enveloppe de Sécurité (F/M)
Core
Design and demonstrate a framework for embedded AI systems to evolve during operation via continual and federated learning while maintaining safety, robustness, and performance within a defined safety envelope.
Role type
Intern, Machine Learning Engineer (Safety-Critical Systems)
Builds
Multi-agent federated learning platform with safety controllers and continual learning mechanisms
Domain
Robotics, Healthcare, Autonomous Drones, Embedded AI
Deliverable
production ML models
Required skills
Python, PyTorch, TensorFlow, C/C++, MATLAB, Machine Learning, Deep Learning, Federated Learning, Continual Learning, Safety Envelope Design
Responsibilities
Define safety envelopes for adaptive AI models, Implement multi-agent federated learning platforms, Integrate continual learning mechanisms, Develop safety controllers to qualify or reject model updates, Evaluate model drift and robustness strategies
