Frame the problem, design the solution, ship it, and own whether it actually moved the metric
Descope to the cheapest version that works, and prototype in days, not weeks
Build for self-service, not for tickets
Treat the platform as a product and internal engineers as your customers
Understand their friction, then solve it with reusable, well-documented tooling
Own and evolve the paved road
Set the direction for our GitOps, Kubernetes (AWS EKS / ArgoCD), and serverless (AWS Lambda) patterns
Own reliability
Lead incident response for the systems you own
Architect AI-driven leverage
Design and operate AI agents and automation
Own access, security, and compliance
Design how we manage credentials and permissions for developers and automated processes
Make things cheaper
Own initiatives that measurably reduce infrastructure cost
Mentor on judgment and systems thinking
Requirements
5+ years building, operating, and owning production infrastructure at scale
Deep expertise in a cloud environment (AWS) and with containers and orchestration (Docker, Kubernetes / EKS, Helm)
Strong command of CI/CD and GitOps (GitHub Actions, ArgoCD) and Git-based workflows
Hands-on expertise with Infrastructure as Code (Terraform) and configuration automation
Solid grasp of Linux (Ubuntu), production databases (PostgreSQL/RDS, DynamoDB, MySQL, MongoDB), and networking (load balancers, firewalls, routing, VPN)
A track record of shipping something you cared about end-to-end and owning the outcome
Treats developer experience as a product — measures adoption and delivery metrics (e.g., DORA) and iterates on tooling the way a PM iterates on a feature
Strong systems-design judgment in ambiguous contexts
Experience leading incidents and operating customer-facing production environments with a focus on uptime and service quality
Uses AI agents and coding tools as a daily driver
Clear written and verbal communication, plus mentorship experience
A security and reliability mindset that fits working with sensitive consumer data
A 4-year degree in engineering or computer science