Cohere
Staff Research Engineer, Model Efficiency
$250K–$535K
New York +3 · Remote · Staff · Full-time · Posted Nov 7, 2025
- Location
- New York
San Francisco
Toronto
Montreal - Education
- PhD in Machine Learning or a related field
About the job
Develops, prototypes, and deploys techniques to improve the speed and efficiency of Cohere’s foundation models in production.
What you'll do
- Develop and deploy improvements to model inference efficiency.
- Work across model architecture, MoE routing, decoding algorithms, and GPU acceleration.
- Optimize performance without compromising model quality.
What they ask for
- PhD in Machine Learning or a related field.
- Understanding of LLM architecture and inference optimization under resource constraints.
- Significant experience with one or more model-efficiency techniques.
- Strong software engineering skills.
Nice to have
- Publications at top-tier conferences or venues such as ICLR, ACL, or NeurIPS.
- Passion for mentoring others.
Pay
CA$250K – CA$535K (base salary)
Tools
LLMs, MoE
AI focus
Improves LLM inference efficiency through architecture and routing optimization, decoding and inference-time algorithms, and GPU software/hardware co-design.
Model-written summary · the original posting is authoritative