Terraform control plane migration
Led adoption of Spacelift as the organization’s Terraform control plane. State lived in git and applies ran from engineer laptops, with no consistent review before production changes landed.
Seven years across AWS and GCP building control planes, paved roads, and network architecture that let engineering teams ship safely by default. Most of my work replaces “someone ran the right command” with a system that converges on its own.
Five problems, and what I did about them.
Led adoption of Spacelift as the organization’s Terraform control plane. State lived in git and applies ran from engineer laptops, with no consistent review before production changes landed.
Built pulpctl, a Python tool that applies a declarative spec to Pulp 3 and turns OS patching into a promotion pipeline. Dev pulls the latest repository snapshot; once it proves out, that exact repo pointer is promoted unchanged to INT, then to PROD.
Designed and implemented a Transit Gateway architecture for a PCI-regulated payments platform, collapsing a mesh of point-to-point VPC connections into a single governed routing domain.
Drove adoption of AI-assisted pull request review across the engineering organization, replacing inconsistent per-team practice with one standard applied to every stack.
Built an observability ecosystem for a distributed payments platform and standardized its data collection on OpenTelemetry, decoupling instrumentation from the vendor tools consuming it.
I’m an infrastructure engineer with over 7 years building and operating production cloud infrastructure across AWS and GCP. I design cloud network architecture, from VPC segmentation through Transit Gateway routing domains, and I build the self-service tooling and Terraform paved roads that let engineering teams provision safely by default.
Most of my work sits on the boundary between operations and software: control planes, reconciliation loops, and guardrails that make the correct thing the easy thing to do.
Currently Cloud Operations Engineer at Gryphon, running a multi-project GCP estate behind enterprise telephony and AI workloads. Based in Westfield, Indiana.

Open to senior infrastructure, platform, and AI infrastructure roles. The fastest way to reach me is email.
andrew@andrewdoesinfra.com