SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 220,000 - 292,000 / annual
Descript is seeking a Senior Software Engineer to own the infrastructure platform that powers the company's operations. This role sits at the foundation of the engineering organization, responsible for compute and deployment, reliability and on-call operations, CI/CD and monorepo health, developer environments, and the infrastructure supporting model training and inference.
You will own multiple critical systems: GCP and Kubernetes infrastructure, Temporal workflows, the GPU fleet behind cloud export, and the deployment and rollback machinery. You'll be part of the on-call rotation and expected to make systems quieter and more actionable. A major focus is building the AI enablement substrate—GPU capacity planning, training and inference pipelines, and ensuring reliability and cost-efficiency as inference scales with user growth.
Cost is treated as an engineering constraint; you'll make smart trade-offs between performance, reliability, and expense. You'll own security boundaries including identity and access, secrets management, least-privilege policies, and supply-chain integrity. The role emphasizes making infrastructure legible through infrastructure-as-code, runbooks, and observability that clearly communicates what failed and why.
You'll work directly with other engineering teams to understand their needs and shape the platform roadmap. Architecture decisions here have lasting impact on a small team covering a large surface area. You'll have real autonomy to decide what to build and what to leave alone. Beyond individual contribution, you'll mentor teammates, provide architectural direction, and raise the team's technical ambition through thoughtful code review and clear communication.
Required: 8+ years building and operating production distributed systems with heavy infrastructure focus. You've run systems where failure was expensive, carried a pager, commanded incidents, and made rollback decisions. You're proficient with major cloud providers and Kubernetes in production, use SLOs and error budgets as operating tools, and default to infrastructure-as-code. You've owned architectures or migrations from planning through launch and can articulate lessons learned. You find unowned work, scope it, earn buy-in, and deliver without a spec.
Experience that strengthens your candidacy: GPU and ML infrastructure (capacity planning, training/inference pipelines, cost and latency optimization), production security engineering (IAM, secrets, supply chain), cloud cost modeling, CI/CD at monorepo scale, video or media workloads, or experience on small platform teams at Series B–D companies.
