Research Engineer / Research Scientist, Post-Training & Reinforcement Learning

What you'll do

• Develop post-training methods, reward models, and multimodal reasoning capabilities, and explore agentic tool use and inference-time optimization.
• Own research from experimental design through distributed training and evaluation, with deployment handoff.
• Create rigorous clinical evaluations and contribute original research and technical insights.

What they ask for

• Expertise in reinforcement learning, language model post-training, or multimodal machine learning.
• Demonstrated ability to develop, train, or improve state-of-the-art machine learning models.
• Strong understanding of post-training techniques, including reinforcement learning, reward modeling, and preference optimization.
• Experience with large-scale distributed training using PyTorch or JAX.

Nice to have

• Publications at leading ML conferences such as NeurIPS, ICML, ICLR, CVPR, or ACL.
• Experience with RL algorithms, reward models or RLHF, large-scale RL infrastructure, multimodal reasoning, inference-time optimization, medical imaging, or reliability evaluation.

Tools

PyTorch, JAX, GRPO, DAPO, vLLM, SGLang, verl, TRL

AI focus

Post-training and reinforcement learning for multimodal medical AI, including reward modeling, clinical reasoning, agentic tool use, and inference-time optimization.