Cohere’s Data Infrastructure team builds the storage and data movement layer that supports model training and evaluation. The team is developing a unified storage system to serve petabytes of training data and model checkpoints across multiple training clusters, with the goal of keeping thousands of GPUs busy. The Staff Software Engineer will architect and lead the design of this system from the ground up, driving its technical vision from concept through production. The role includes deploying and operating the platform across Kubernetes clusters at petabyte scale, defining organizational standards and best practices, and working with researchers and training-infrastructure teams to translate workload patterns into platform requirements. The engineer will address networking, I/O, and consistency challenges involved in moving large datasets and checkpoints across regions and backends, with GPU idle time and time-to-insight as key success measures.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
Job Type
Full-time
Career Level
Senior
Education Level
No Education Listed