ByteDance
Software Engineer - Compute Infrastructure (Orchestration & Scheduling)
Seattle, Washington · Full-time · Posted Oct 28, 2026
- Experience
- 2+ years of relevant industry experience; new graduates with a Ph.D. degree and strong publication records can be an exception.
- Education
- B.S./M.S. degree in Computer Science, Computer Engineering, or a related area; new graduates with a Ph.D. degree and strong publication records can be an exception.
About the job
Build and improve large-scale compute infrastructure, including Kubernetes-based cluster management, unified workload scheduling, and platforms for AI and LLM training and inference.
What you'll do
- Enhance Kubernetes-based cluster platforms for performance, scalability, and resilience.
- Design and maintain unified scheduling for diverse workloads and heterogeneous resources.
- Develop AI-model-based scheduling to improve workload performance and resource utilization across global data centers.
What they ask for
- B.S./M.S. degree in Computer Science, Computer Engineering, or a related area; new Ph.D. graduates with strong publication records may be an exception.
- At least 2 years of relevant industry experience, with the stated Ph.D. exception for new graduates.
- Solid understanding of at least one listed area: Unix/Linux, distributed and parallel systems, high-performance networking, or large-scale software systems; experience building cloud or ML infrastructure.
- Familiarity with Docker and Kubernetes, and proficiency in at least one of Python, Go, C++, Rust, or Java.
Nice to have
- Experience with large-scale cluster management systems such as Kubernetes, Ray, Yarn, or Mesos, or with resource efficiency and job scheduling.
- Experience with application scaling, workload co-location, isolation, or public cloud and ML services.
Tools
Kubernetes, Docker, Ray, Yarn, Mesos, AWS, Azure, GCP
AI focus
Use AI models to optimize workload performance and resource utilization; develop compute platforms for ML and LLM training and inference.
Model-written summary · the original posting is authoritative