Senior Infrastructure Engineer – Cloud Development Environments

Requirements

• 8+ years in infrastructure, platform, or DevOps engineering roles.
• Deep AWS expertise: networking, compute (EC2, EKS, Fargate), IAM, and cost management.
• Proficiency with Terraform and modern IaC practices (modules, remote state, policy-as-code).
• Hands-on experience with Coder, Gitpod, DevPod, or equivalent CDE/remote development platforms.
• Strong container and Kubernetes knowledge: scheduling, resource management, node pools, and storage.
• Solid understanding of observability principles and experience with Prometheus/Grafana or equivalent stacks.
• Security-first attitude with experience implementing zero-trust networking and secrets management (e.g. Vault, AWS Secrets Manager).

What You Will Be Doing

• Architect and operate the Coder-based CDE platform on AWS, ensuring high availability, performance, and security for 1,000+ concurrent users and agents.
• Define and manage Coder workspace templates (Terraform) for diverse workloads: frontend, backend, ML, and autonomous AI agents.
• Own infrastructure-as-code (IaC) using Terraform and manage state backends, module libraries, and CI/CD pipelines for infra changes.
• Build and maintain golden workspace images optimized for fast cold-start and minimal drift, using versioned base images and automated build pipelines.
• Continuously improve developer productivity metrics: workspace start time, environment parity, and time-to-first-commit for new hires, treating internal engineers as customers and drive SLAs, change logs, and feedback loops.
• Tune IDE performance within Coder workspaces, managing extension marketplaces, LSP servers, and workspace resource profiles.
• Develop and sustain a comprehensive observability stack (metrics, logs, traces) for the CDE platform using tools such as Prometheus, Grafana, OpenTelemetry, and AWS CloudWatch, define and track SLOs/SLAs, and run blameless post-mortems on incidents.
• Implement alerting, runbooks, and on-call rotation for the platform, targeting < 5-minute MTTD and < 30-minute MTTR for P1 incidents.
• Design workspace profiles and resource tiers tailored for AI coding agents (e.g. Claude, GitHub Copilot Workspace, custom agents), including GPU access, model API egress, and sandboxed execution environments.
• Stay ahead of with the rapidly evolving AI developer tooling landscape and evaluate new CDE and agentic workflow integrations.