Research Engineer Graduate (AI Training Systems & RL Infrastructure - Seed Infra) - 2026 Start (PhD)

ByteDance is a pioneering tech company dedicated to advancing artificial general intelligence. The Research Engineer Graduate will conduct R&D on AI infrastructure and support the training of foundation models, while collaborating with a team to develop scalable solutions.

What you'll do:

• Conduct research and development on large-scale AI infrastructure to support efficient training and post-training of foundation models, multimodal LLMs, and image/video generation models
• Design and optimize distributed training strategies, including data/model/tensor/pipeline/expert parallelism, computation–communication overlap, and large-scale GPU cluster scaling
• Prototype and improve end-to-end reinforcement learning (RL) training systems, covering rollout generation, policy optimization, evaluation, and iterative deployment workflows
• Build scalable and fault-tolerant infrastructure that operates reliably under dynamic workloads and heterogeneous compute environments
• Analyze performance bottlenecks across the training stack (e.g., networking, scheduling, GPU memory management), and develop principled optimization approaches to improve throughput, efficiency, and stability
• Develop tooling, monitoring, debugging, and observability frameworks to ensure reliability of large-scale training and RL systems
• Collaborate with researchers and engineers on system–algorithm co-design, translating research prototypes into scalable, production-ready infrastructure systems

What they're looking for:

• Individuals who are completing or have recently completed a PhD in Computer Science, Electrical Engineering, or a related technical field (graduating students welcome)
• Strong background in distributed systems, large-scale machine learning systems, or deep learning infrastructure
• Research or hands-on experience in training or optimizing large-scale models (e.g., LLMs, multimodal models, RL systems)
• Understanding of parallelism strategies (e.g., data, model/tensor, pipeline, expert parallelism) and distributed training concepts
• Familiarity with reinforcement learning workflows such as rollout generation, policy optimization, and evaluation loops
• Proficiency in programming (e.g., Python and/or C++) and experience with modern ML frameworks (e.g., PyTorch and distributed training tools)

Benefits:

• Day one access to medical, dental, and vision insurance
• A 401(k) savings plan with company match
• Paid parental leave
• Short-term and long-term disability coverage
• Life insurance
• Wellbeing benefits, among others
• 10 paid holidays per year
• 10 paid sick days per year
• 17 days of Paid Personal Time (prorated upon hire with increasing accruals by tenure)
• Employees also receive 10 paid holidays per year, 10 paid sick days per year and 17 days of Paid Personal Time (prorated upon hire with increasing accruals by tenure).