Remote Staff Engineer, RL & Integration

Cohere; Greater London; Remote; GBP 110,000 - 150,000; Full time

Benefits offered by this job

Weekly lunch stipend

Health and dental benefits

Parental leave

Job summary

Cohere is a leading security-first enterprise AI company with a London presence. We are expanding the integration team to scale machine learning algorithms and infrastructure for post-training of large models.

You will design scalable software for training, collaborate with cross-disciplinary teams, and contribute to production-grade post-training systems. This remote-friendly role accepts candidates across UTC−06:00 to UTC+01:00, emphasizing collaboration, speed, and high-performance ML

Qualifications

• Strong software engineering skills with production-grade experience.
• Experience with ML frameworks and large-scale model training.
• Familiarity with distributed training and performance profiling.

Responsibilities

• Design and write high-performing, scalable software for training models.
• Develop tools to accelerate research and LL M training.
• Coordinate with Infrastructure, Efficiency, Serving, and scientific teams to build a cohesive post-training ecosystem.
• Implement techniques to improve performance and speed of training cycles (SFT, offline, RL).
• Research and experiment with cluster and data infrastructure.

Skills

Software engineering

Python

ML frameworks (PyTorch/JAX)

Distributed training

Test-driven development

Education

Bachelor’s degree in CS/Math/Engineering

Tools

Kubernetes

Ray

Docker

Job description

Cohere is a leading security-first enterprise AI company with a London presence. We are expanding the integration team to scale machine learning algorithms and infrastructure for post-training of large models.

You will design scalable software for training, collaborate with cross-disciplinary teams, and contribute to production-grade post-training systems. This remote-friendly role accepts candidates across UTC−06:00 to UTC+01:00, emphasizing collaboration, speed, and high-performance ML