GPU Fleet SRE for AI Inference Platform

ApplyMint; New York, San Francisco (NY, CA); On-site; USD 180,000 - 270,000; Full time

Benefits offered by this job

Competitive salary

Meaningful equity

Health, dental, and vision benefits

Cloud native community events

Fitness stipend and learning budget

Job summary

Beam is an ultrafast AI inference platform building a serverless runtime that launches GPU-backed containers in under a second and scales to thousands of GPUs.

The role centers on owning compute fleet health end to end, automating failure triage, and designing a GPU qualification platform with burn-in and baselining for new hardware. You will also own firmware telemetry and low-level log collection critical to repair automation.

Qualifications

• Strong understanding of firmware and silicon failure modes.
• Fluent with AI tooling and maximizing token usage when appropriate.
• Experience debugging production incidents from triage to post-mortem.
• Enthusiasm for developer tools and cloud-native open source software.

Responsibilities

• Own compute fleet health end to end, building metrics pipelines, alerting, and a unified health view.
• Turn deployment debugging into an automated pipeline, from detection to triage.
• Design the GPU qualification platform for burn-in, baselining, and NPI on new GPUs.
• Own firmware telemetry, log collection at scale, and the low-level access layer for repair tooling.

Skills

Hardware intuition

AI tooling fluency

Production debugging

Cloud native & open source

Job description

Beam is an ultrafast AI inference platform building a serverless runtime that launches GPU-backed containers in under a second and scales to thousands of GPUs.

The role centers on owning compute fleet health end to end, automating failure triage, and designing a GPU qualification platform with burn-in and baselining for new hardware. You will also own firmware telemetry and low-level log collection critical to repair automation.