AI Foundation Infra Intern: Distributed Training & Inference

ByteDance; Seattle (WA); On-site; USD 36,000 - 58,000; Full time

Job summary

ByteDance Seed in Seattle is seeking an Infrastructure Intern to help build and optimize large-scale distributed training systems, improve RL pipelines, and enhance inference performance in a cutting-edge AI infrastructure environment.

You will collaborate with researchers and engineers to translate model requirements into scalable system solutions, and gain hands-on experience in ML systems, compiler/runtime optimizations for accelerators, and performance profiling.

Qualifications

• Pursuing a Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, or related fields.
• Proficiency in Python or C++
• Familiarity with PyTorch or similar ML frameworks
• Strong analytical and problem-solving skills
• Ability to collaborate in a fast-paced environment
• Interest in ML systems or AI infrastructure

Responsibilities

• Assist in building and optimizing large-scale distributed training systems.
• Support development of reinforcement learning training pipelines.
• Improve inference performance, including latency and throughput.
• Contribute to compiler or runtime optimizations for accelerators.
• Conduct performance analysis, profiling, and bottleneck identification.
• Develop internal tools to improve infrastructure efficiency.
• Collaborate with researchers and engineers to translate model requirements into scalable solutions.

Skills

Python

C++

PyTorch

Education

Bachelor’s or Master’s degree in CS/EE or related

Job description

ByteDance Seed in Seattle is seeking an Infrastructure Intern to help build and optimize large-scale distributed training systems, improve RL pipelines, and enhance inference performance in a cutting-edge AI infrastructure environment.

You will collaborate with researchers and engineers to translate model requirements into scalable system solutions, and gain hands-on experience in ML systems, compiler/runtime optimizations for accelerators, and performance profiling.