Software Engineer, ML Workload Platform
New
C
CohereAI infrastructure
Location: Canada; Secondary Locations: United States; Workplace: RemoteFull-Time
Salary$175K - $325K; $175K – $325K • Offers Equity • Multiple Ranges; USA - California, New York and Washington: Base Salary $200K – $325K • Offers Equity; USA - All Other States: Base Salary $175K – $290K • Offers Equity; Canada: Base Salary CA$250K – CA$400K • Offers Equity
Apply NowOpens the employer's application page
Job Details
- Required Skills
- PythonKubernetesGoDistributed Systems
Requirements
- Have run a job platform, batch system, or Kubernetes controllers in production.
- Have experience with scheduling, queues, and multi-tenant fairness.
- Demonstrate distributed systems judgment, including handling retries and partial failure.
- Have worked on ML training workflows, RL sandboxes, container networking (CNI), or checkpoint storage at scale.
- Have written software in Python or Go.
- Have worked with researchers, ML engineers, or teams with differing goals.
- Have identified a bottleneck, proposed a fix, and driven the work in a fast-moving environment.
- Direct GPU-cluster experience is a plus.
Responsibilities
- Own the Kubernetes platform that schedules and runs training, evaluation, and reinforcement-learning jobs.
- Manage job capacity, failure handling, and visibility into whether platform issues caused job failures.
- Build networking and runtime isolation for workloads, including agent and sandbox environments.
- Work with the data infrastructure team on storage for checkpoints and other training artifacts.
- Troubleshoot production failures and performance problems, and improve system operability.
- Build automation, infrastructure-as-code, and observability for the platform.
View Full Description & ApplyYou'll be redirected to the employer's site
