Senior ML Platform Engineer: Distributed Training & Scheduling

ByteDance; Seattle (WA); On-site; USD 207,000 - 368,000; Full time

Job summary

ByteDance in Seattle seeks a senior distributed ML systems engineer to optimize scheduling, build scalable training/inference runtimes, and drive online orchestration for next-gen recommender models. You will work across Kubernetes-based frameworks and large-scale ML workloads, contributing to both research and production platforms.

The role requires 5+ years in Go or Python, strong distributed systems knowledge, and ability to design robust, scalable engineering solutions in a fast-paced team

Qualifications

• Bachelor's degree or above in Computer Science or similar field.
• At least 5 years of experience with Go or Python in a Linux environment.
• Familiar with distributed scheduling frameworks and ML systems.
• Master the principles of distributed systems and design/maintain large-scale systems.
• Strong analytical, documentation, and self-motivation skills.

Responsibilities

• Optimize resource efficiency in distributed orchestration and scheduling.
• Develop scheduling frameworks around Kubernetes/Godel ecosystem for various scenarios.
• Extend AutoScaling and automatic parallelization for models and operations.
• Manage preemption/eviction, cross-cluster resource docking, and multi-datacenter runtime.
• Build training system architecture for ultra-large recommendation models.
• Design distributed training runtimes and ML model synchronization.
• Interface with platform to improve diagnosability of distributed training.
• Construct online orchestration for next-gen Recommender system and inference.

Skills

Go

Python

Distributed systems knowledge

Strong coding skills

Education

Bachelor's degree in Computer Science or similar

Tools

Kubernetes

Yarn

Flink

MapReduce

Mesos

Celery

Job description

ByteDance in Seattle seeks a senior distributed ML systems engineer to optimize scheduling, build scalable training/inference runtimes, and drive online orchestration for next-gen recommender models. You will work across Kubernetes-based frameworks and large-scale ML workloads, contributing to both research and production platforms.

The role requires 5+ years in Go or Python, strong distributed systems knowledge, and ability to design robust, scalable engineering solutions in a fast-paced team