Summer 2027 PhD AI Research Infrastructure, RL Post-Training Intern
Core
Develop and optimize infrastructure for reinforcement learning (RL)-based post-training of large language and multimodal models.
Role type
PhD research intern (AI infrastructure)
Builds
Scalable systems for rollout generation, inference, reward computation, and policy updates
Domain
AI research, distributed systems, large-scale model training
Deliverable
production ML models
Required skills
Python, PyTorch, reinforcement learning, LLM post-training, distributed training, multi-GPU workloads, parallelism strategies, GPU performance, memory management, networking, system profiling
Preferred skills
Containerization, experiment tracking, cloud/cluster computing, publications at top venues (ICML, NeurIPS, etc.)
Technologies
PyTorch, Python, multi-GPU clusters
Responsibilities
Develop RL post-training infrastructure, build scalable rollout/inference systems, optimize distributed training efficiency, design researcher interfaces, build experiment tools, profile training pipelines, support on-policy/off-policy workflows, document system designs
Seniority
PhD student, research intern
