Build Kubernetes training infrastructure at Waabi, using Python and AWS to support Physical AI for autonomous trucks and robotaxis. This senior engineering role develops platforms used by ML researchers and engineers. You will shape APIs, CLIs, experiment workflows, and reliable production systems. The work spans distributed PyTorch training, GPU scheduling, data pipelines, and model delivery. Key Responsibilities:
Build Kubernetes infrastructure for distributed GPU training workloads Design developer-facing CLIs, SDKs, templates, and workflows Strengthen dataset versioning, sharding, and artifact loading Implement observability for utilization, failures, queues, and costs Ship CI/CD for models alongside autonomy and simulation Requirements:
Five or more years in software or infrastructure engineering Excellent Python and experience designing usable APIs and CLIs Practical AWS, IAM, GPU, networking, and infrastructure-as-code expertise Distributed PyTorch training with experiment tracking and model registries Fluency with containers, CI/CD, and up-to-date build systems Use Python, AWS, Kubernetes, and distributed training expertise to advance Waabi's autonomous transportation infrastructure.