Research Scientist, Discovery (Engineering & Applied Sciences)
Core
Design and build reinforcement learning environments, rewards, and training data to improve Claude's capabilities in engineering and applied sciences.
Role type
Senior IC machine learning researcher (AI for science)
Builds
Training environments, evaluation frameworks, and simulation-based training runs for frontier models
Domain
Artificial Intelligence, Machine Learning, Engineering & Applied Sciences
Deliverable
production ML models
Required skills
Python, PyTorch or JAX, training/evaluating ML models, hypothesis formation, experimental design, complex system debugging, rapid domain learning
Preferred skills
Reinforcement learning, pre-training on LLMs, environment/reward design, large-scale LLM training pipelines, AI for science, long-horizon reasoning, simulator/surrogate model building, large-scale data pipelines
Technologies
Python, PyTorch, JAX
Responsibilities
Design reinforcement learning environments and rewards; Create benchmarks for engineering workflows; Develop simulation-based training environments; Research long-horizon task completion and multi-agent collaboration; Scale research ideas to production training runs
Seniority
Senior, hands-on IC
