Graduate ML Systems Scientist - LLM Training & Inference

ByteDance; San Jose (CA); On-site; USD 162,000 - 317,000; Full time

Benefits offered by this job

Medical, dental, vision insurance

401(k) with company match

Parental leave

Job summary

ByteDance in the United States seeks a PhD-level engineer to develop and optimize large-scale LLM training and inference systems. You will work on GPU/CUDA performance, collaborating with researchers to push LLM training and RL forward on massive distributed architectures.

You’ll join a global team spanning the US, China and Singapore, contributing to scalable, reliable ML infrastructure for a broad portfolio of products.

Qualifications

• PhD in Software Development, Computer Science, Computer Engineering, or related field completed or near completion.
• Strong grasp of algorithms and data structures; proficient in Python.
• Familiar with deep learning basics, neural networks, and PyTorch.

Responsibilities

• Develop and optimize LLM training, inference, and RL frameworks.
• Collaborate with model researchers to scale LLM training and RL.
• Optimize GPU/CUDA performance for high‑performance LLM workloads.

Skills

PhD in Computer Science/Software Eng

Python

Algorithms

Data Structures

Education

PhD in CS/CE/SE

Tools

CUDA

PyTorch

Megatron-LM

TensorRT-LLM

JAX/TPU

Job description

ByteDance in the United States seeks a PhD-level engineer to develop and optimize large-scale LLM training and inference systems. You will work on GPU/CUDA performance, collaborating with researchers to push LLM training and RL forward on massive distributed architectures.

You’ll join a global team spanning the US, China and Singapore, contributing to scalable, reliable ML infrastructure for a broad portfolio of products.