Krea AI

The role focuses on building and operating the infrastructure for AI research and inference, including distributed training, large GPU clusters, petabyte-scale data pipelines, custom distributed datastores, job orchestration systems, and streaming pipelines. It involves designing multi-stage data pipelines, managing distributed training and inference on large GPU clusters, scaling workloads, and optimizing dataloaders and networking for large training runs. The role requires strong systems thinking and experience with distributed systems, Python, Kubernetes, and data tools.