Graduate ML Engineer: Orchestration & Online Model Serving

ByteDance; San Jose (CA); On-site; USD 180,000 - 230,000; Full time

Job summary

ByteDance is building a Data-AML-Engine Orchestration platform to power online model serving across TikTok and other ByteDance products. You will contribute to orchestration, scheduling, and resource management systems connecting heterogeneous compute with production ML workloads.

Join a team focused on scalable infrastructure—enabling low latency, high availability, and efficient GPU usage while tackling challenging distributed systems problems across multiple clusters and services.

Qualifications

• Bachelor's or Master's degree in Computer Science, Software Engineering, AI, or related field.
• Proficiency in Go, C++, or Python with strong CS fundamentals.
• Familiarity with Linux, OS concepts, networks and distributed systems.
• Strong hands-on and exploratory abilities with code, metrics, logs, and profiling.

Responsibilities

• Design and build foundational orchestration capabilities for ML platforms (Kubernetes Operators, containers, jobs/stateful workloads).
• Create multi-tenant resource and quota systems; improve GPU utilization and cost efficiency via resource pooling and FinOps.
• Develop lifecycle orchestration for online model serving (deployment, upgrades, autoscaling, disaster recovery).
• Build serving orchestration and traffic management for disaggregated clusters (topology-aware scheduling, KV Cache affinity, QoS/SLA).

Skills

Go

C++

Python

Distributed systems

Linux

Data structures

Education

Bachelor's or Master's in Computer Science, Software Engineering, or related field

Tools

Kubernetes

Container runtimes

Volcano

Koordinator

OpenKruise

Model serving systems

Job description

ByteDance is building a Data-AML-Engine Orchestration platform to power online model serving across TikTok and other ByteDance products. You will contribute to orchestration, scheduling, and resource management systems connecting heterogeneous compute with production ML workloads.

Join a team focused on scalable infrastructure—enabling low latency, high availability, and efficient GPU usage while tackling challenging distributed systems problems across multiple clusters and services.