Member of Technical Staff - Infrastructure

The role

You will build systems that turn heterogeneous accelerator hardware into reliable production infrastructure for Gimlet's AI cloud. You will work across bare metal, Linux, Kubernetes, cluster scheduling, observability, and automation.

What you will do

• Deploy and operate production clusters across different accelerator architectures
• Automate hardware provisioning, validation, upgrades, and fleet lifecycle management
• Improve cluster scheduling, resource utilization, isolation, and capacity management
• Build observable infrastructure that enables faster debugging, incident response, and recovery
• Partner across distributed systems, runtime, compiler, networking, and hardware teams to bring new accelerators into production

Required qualifications

• Experience in infrastructure, cluster engineering, platform engineering, SRE, or HPC
• Strong Linux systems knowledge and production debugging experience
• Experience operating Kubernetes, Slurm, Nomad, or similar orchestration systems
• Experience automating infrastructure with Python, Go, Terraform, Ansible, or similar tools
• Experience with GPU or accelerator infrastructure, including drivers, firmware, or CUDA/ROCm
• The ability to build systems that are observable, recoverable, and reliable in production
• A bachelor’s degree in a relevant field or equivalent practical experience

Nice to have

• Experience building or operating AI inference, training, HPC, or neocloud infrastructure
• Experience with bare-metal provisioning, PXE/iPXE, image pipelines, BIOS/firmware management, or rack bring-up
• Experience with multi-tenant cluster isolation, quota systems, fair scheduling, or usage accounting
• Experience debugging distributed workload performance across compute, memory, network, and storage bottlenecks
• Experience building observability platforms using technologies such as Prometheus, OpenTelemetry, Grafana, or similar tooling
• Familiarity with heterogeneous hardware environments across NVIDIA, AMD, Intel, ARM, or emerging accelerators

About the company

Gimlet is building the first multi-silicon neocloud designed for fast, efficient AI inference. They combine large-scale compute infrastructure with an execution platform that partitions AI workloads and maps each stage to the hardware best suited to run it.