Senior AI Infrastructure Platform Engineer

descript; United States; On-site; USD 220,000 - 292,000; Full time

Benefits offered by this job

Equity

Competitive benefits

Job summary

descript owns the platform that the company runs on: compute and deployment, reliability and on-call, CI/CD and monorepo health, and the infrastructure that model training and inference run on, with security boundaries around it.

You will contribute to how models are trained and served here, as well as how agents work inside our codebase: the environments they run in, the verification that makes their output trustworthy, and the review paths that keep it all legible.

Qualifications

• 8+ years building and operating production distributed systems or equivalent server-side engineering with heavy infrastructure focus.
• Experience leveraging agents to multiply impact and integrating AI into workflows.
• Experience with reliability, deployment decisions, and incident management.
• Experience carrying a pager, handling incidents, and performing rollbacks.

Responsibilities

• Own the platform including GCP, Kubernetes, Temporal, and GPU fleet infrastructure.
• Own AI enablement substrate: capacity, training and inference pipelines, and production serving costs.
• Balance cost with performance; apply metering and attribution behind usage.
• Ensure security with identity, secrets management, and least-privilege controls.
• Write infrastructure-as-code and maintain runbooks and in-repo context for onboarding.

Skills

Distributed systems

Production deployment

Incident response

SLOs and error budgets

Kubernetes in production

Infrastructure as code

Cloud platforms

Tools

Kubernetes

Terraform

IaC

Job description

descript owns the platform that the company runs on: compute and deployment, reliability and on-call, CI/CD and monorepo health, and the infrastructure that model training and inference run on, with security boundaries around it.

You will contribute to how models are trained and served here, as well as how agents work inside our codebase: the environments they run in, the verification that makes their output trustworthy, and the review paths that keep it all legible.