Senior AI Infrastructure Engineer

The role

You will design, build, and scale a high-performance AI platform for autonomous driving models. This role involves managing distributed training, experiment tracking, and model deployment to bridge the gap between research and production.

What you will do

• Scale research workloads using PyTorch Distributed and Ray Train
• Architect and optimize multi-GPU setups for model and data parallelism
• Optimize low-level communication like NCCL, InfiniBand, or RoCE v2
• Optimize hardware utilization through Kubernetes-native GPU scheduling
• Deploy and scale optimized model artifacts using TensorRT, ONNX Runtime, and Triton Inference Server
• Architect and deploy Autonomous AI Agents to monitor GPU cluster health
• Develop agent-driven automation for infrastructure code and resource optimization
• Support researchers in building Data Machines using AI agents for data curation
• Design and maintain ML infrastructure leveraging MLFlow, Argo Workflows, and Kubernetes
• Implement deployment strategies including A/B testing, shadow deployments, and rollback mechanisms

Required qualifications

• 5+ years in ML infrastructure, MLOps, or DevOps supporting high-scale compute environments
• Deep understanding of multi-GPU training strategies (FSDP, DeepSpeed, Ray Train)
• High-performance networking (NCCL, InfiniBand)
• Mastery of Kubernetes, Terraform, and Helm
• Proven experience building or supporting Agentic Workflows for infrastructure or data automation
• Expertise in MLFlow, Argo Workflows, and Kubernetes
• Strong experience with Docker, Kubernetes, and Helm
• Proficiency in Apache Airflow, Kafka, Spark, and GitOps automation
• Proficiency in Python and Bash

Nice to have

• Experience with Go or Rust
• Familiarity with the Model Context Protocol (MCP)
• Experience in hybrid cloud and on-prem GPU cluster management for Physical AI workloads
• Experience utilizing LLMs for semantic monitoring and log analysis

About the company

Gatik is a leader in autonomous middle-mile logistics providing autonomous transportation-as-a-service. The company uses proprietary Level 4 autonomous technology to transport freight between pick-up and drop-off locations.