Senior Cloud Infrastructure Engineer

Who we are Gatik, the leader in autonomous middle-mile logistics, is revolutionizing the B2B supply chain with its autonomous transportation-as-a-service (ATaaS) solution and prioritizing safe, consistent deliveries while streamlining freight movement by reducing congestion. The company focuses on short-haul, B2B logistics for Fortune 500 retailers and in 2021 launched the world’s first fully driverless commercial transportation service with Walmart. Gatik's Class 3-7 autonomous trucks are commercially deployed across major markets, including Texas, Arkansas, and Ontario, Canada, driving innovation in freight transportation. The company's proprietary Level 4 autonomous technology, Gatik Carrier™, is custom-built to transport freight safely and efficiently between pick-up and drop-off locations on the middle mile. With robust capabilities in both highway and urban environments, Gatik Carrier™ serves as an all-encompassing solution that integrates advanced software and hardware powering the fleet, facilitating effortless integration into customers' logistics operations. About the role We are seeking a Senior Cloud Infrastructure Engineer to architect and manage the large-scale compute and data infrastructure powering our autonomous driving stack. While researchers develop perception, planning, and world models, your mission is to build the high-performance systems and pipelin [... source excerpt omitted ...] ters, distributed training frameworks, and automated workflows are scalable, resilient, and cost-effective. This role is onsite 5 days a week at our Santa Clara, CA office! What you'll do Cloud-Native Orchestration & Kubernetes Advanced K8s Management: Architect and maintain mission-critical Kubernetes clusters optimized for heavy GPU/TPU workloads. GPU Scheduling: Implement and optimize Kubernetes-native GPU scheduling (NVIDIA GPU Operator) to ensure maximum hardware utilization. Infrastructure as Code: Drive the "Everything as Code" philosophy using Terraform, Helm, and cloud-native tools. Self-Healing Infrastructure: Deploy Autonomous AI Agents (LangGraph, CrewAI) to [... source excerpt omitted ...] l lifecycles using Airflow and Kubernetes to streamline the transition from training to simulation. High-Performance Serving: Support the deployment of models into simulation and production environments using Triton Inference Server, Ray Serve, and ONNX Runtime. Distributed Training & ML Systems Support Training Systems Support: Enable researchers to scale models (VLA, World Models) across multi-node setups using PyTorch Distributed (TorchElastic), Ray Train, and Horovod. Networking Optimization: Optimize low-level communication (e.g., NCCL tuning, InfiniBand, or RoCE v2) to minimize latency for 3D Gaussian Splatting (3DGS) and large-scale training. Hardware-Aware Orchestratio