I run AWS at scale, and make it cheaper to run.

Scaling a cost scheduler across the AWS organization

Non-production EC2, ECS and RDS ran around the clock across dozens of accounts. I took a scheduler that covered 8 accounts and rebuilt it as a reusable Terraform module, with Lambda doing the start/stop work and GitHub Actions rolling it out account by account. It now covers 50+ accounts.

I've rebuilt the pattern from scratch as an open-source module: terraform-aws-resource-scheduler.

Hybrid Fargate Spot service for Tableau Bridge

Designed an ECS module that mixes Fargate Spot and On-Demand capacity, with memory-based auto-scaling and custom Docker images for PostgreSQL compatibility, so the service stays available while most of it runs on Spot pricing.

Importing legacy infrastructure into Terraform

Wrote Python automation to import hand-built AWS resources across 30+ accounts into Terraform state, recovered a lost Azure state file, and moved PagerDuty alerting and on-call configuration into Terraform so monitoring follows the same review process as infrastructure.

Multi-account provisioning at Amazon

Automated account provisioning for enterprise customers with Control Tower and Account Factory for Terraform, so governance guardrails apply the same way every time. Handled 2–4 SEV-1/SEV-2 escalations a week on Control Tower, landing zones and Service Control Policies for organizations of 100+ accounts.

Enterprise SSO when the owner left

Took over client SSO integrations after a key departure and resolved AD FS and SAML failures for enterprise customers including Amazon, Boston Scientific and Cardinal Health.

A CloudOps AI framework

Building an internal framework of agent definitions, prompts and per-repository configuration, so engineers can start a coding agent that already knows each infrastructure repo's structure and rules. Connected to AWS, GitHub and Atlassian through MCP for resource imports, PR review and Jira triage.