Staff Engineer, Training Infrastructure & ML Pipelines

Cohere; United States; Remote; USD 140,000 - 200,000; Full time

Benefits offered by this job

Lunch stipend

Health and dental benefits

401K / Pension

Parental leave (6 months)

Enrichment benefits + conferences

6 weeks paid vacation

Offsite travel budget

Home office stipend

Job summary

Cohere is hiring for a role focused on training pipelines, productionizing state-of-the-art models, and bridging research with production systems. You will work across software and research boundaries, leveraging compute, data, and talent to push model performance and reliability.

You will design and refine training infrastructure, write production-grade code, and collaborate with top researchers to deploy scalable AI solutions.

Qualifications

• Advanced software engineering skills and production code experience.
• Strong Python programming and ML framework experience (JAX, PyTorch, XLA/MLIR).
• Experience with distributed training infrastructures (Kubernetes, Slurm) and related tools (Ray).
• Hands-on experience training large models at scale and building training infra tooling.

Responsibilities

• Design high-perf scalable training software.
• Improve training infrastructure performance and scale.
• Build tools to speed up training cycles and improve data infra.
• Research and experiment with supercompute and data infra ideas.
• Collaborate with leading researchers and engineers.

Skills

Python programming

ML frameworks

Distributed training

Kubernetes

Research collaboration

Education

PhD or MS in CS or related field

Tools

JAX

PyTorch

XLA/MLIR

Job description

Cohere is hiring for a role focused on training pipelines, productionizing state-of-the-art models, and bridging research with production systems. You will work across software and research boundaries, leveraging compute, data, and talent to push model performance and reliability.

You will design and refine training infrastructure, write production-grade code, and collaborate with top researchers to deploy scalable AI solutions.