Reinforcement Learning Engineer
amgen
Job Description
- Partner with data scientists and software engineers to frame business problems as reinforcement learning and identify suitable approaches.
- Implement and evaluate reinforcement learning methods, such as contextual bandits, policy learning, or preference-based approaches, with technical guidance, for applications that learn from feedback.
- Build and scale Python-based ML systems from prototypes to production using frameworks such as PyTorch, scikit-learn, or other relevant machine learning libraries.
- Design evaluation methods and analyze results using appropriate metrics, baselines, statistical methods, and offline or online experiments.
- Work with human and AI-generated feedback, evaluation data, and interaction traces to support model improvement and learning from preferences.
- Implement DevOps/MLOps best practices to build reliable, reproducible ML workflows, including experiment tracking, data and model versioning, testing, and deployment.
- Collaborate with product and platform teams to integrate models into AI applications and monitor their performance after deployment.
- Document technical work, communicate findings, and keep current with relevant developments in reinforcement learning and machine learning.