Tech Stack

Description

You will own the end-to-end strategy for platform reliability, focusing on GPU/TPU infrastructure and workload orchestration. You'll design robust test harnesses, optimize inference services, and overhaul monitoring systems to ensure research workflows run smoothly.

Requirements

• Proven experience in architecting and managing large-scale AI/ML workloads in production
• Expertise in Google Cloud Platform (GCP) cloud compute design
• Significant experience deploying and managing complex workloads on Kubernetes (GKE)
• Professional familiarity with NVIDIA GPU generations and high-performance compute
• Strong programming skills with a reliability-first approach

Responsibilities

• Own end-to-end strategy for platform reliability focused on accelerator infrastructure and workload orchestration
• Design and implement a robust test harness to validate infrastructure upgrades
• Architect and optimize next-generation inference services for high-throughput performance
• Overhaul logging and monitoring systems for proactive alerting and telemetry
• Improve internal CI/CD stability and reduce failure rates