Lead a 7-person DevOps team running AWS (EKS) and Azure (AKS) platforms for every product team in the company. Replaced a patchwork of tools with one uniform delivery solution and built the AI layer the team runs on.

• Took the organization from an assortment of tools, pipelines and manual deploys to one uniform delivery solution across AWS and Azure: all infrastructure in Terraform, build and test in GitHub Actions, and deployment through Argo CD GitOps.
• Retired Octopus Deploy and TeamCity in favor of GitHub Actions and Argo CD, using an AI-assisted migration skill that converts each service's build pipeline.
• Brought all infrastructure across every AWS account and Azure subscription under Terraform, with coverage reported weekly to department heads.
• Made Argo CD the company-standard GitOps engine and retired Flux. Piloted a hub-and-spoke layout with argocd-agent and fixed a server-side-apply diff problem that was blocking delivery from the hub.
• Built new platforms on the standard pattern: a product platform across four AWS environments (self-hosted GitHub Actions runners on EKS, Datadog, private ingress, per-account secrets) and Temporal on AWS and Azure with mTLS.
• Automated Kubernetes upgrades for EKS, AKS and Argo CD (cluster discovery, version-path analysis, preflight audit, dry run) behind a browser-based upgrade command center. Migrated staging and production from RDS to Aurora with a reusable scripted cutover.

• Built an in-house AI tier 1 on-call assistant that is the first line of defense on S1 alerts. With access to FireHydrant, Jira, Confluence and the team knowledge base, it investigates the alert, researches what is going on, and gives the on-call engineer a proposed fix. The engineer decides and acts.
• Built the team brain into Claude Code as a shared plugin: architecture, runbooks and conventions load into every engineer's session, so engineers become subject-matter experts much faster and can respond to issues across every team in the company.
• Designed an agentic engineering-management layer on Claude Code: custom skills and scheduled headless agents that pull from Jira, Datadog, Outlook and Teams to run sprint planning, backlog grooming, help desk triage and routing, on-call review and weekly ops reporting.
• Enforced AI guardrails as policy-as-code: hooks shipped with the team plugin stop agents from making non-compliant commits and PRs or closing tickets without a root cause.
• Used agents with persistent memory to root-cause long-standing production issues: an N+1 call fan-out behind recurring 504s, shared-library memory leaks causing out-of-memory kills, and duplicate private DNS zones behind an outage, found by auditing DNS across every AKS cluster.
• Scoped an AI observability platform (Langfuse on dedicated AWS accounts and EKS) and set a multi-agent coding pattern where Opus orchestrates and Sonnet subagents handle routine work, to keep model cost down.

• Sponsored two engineers for senior promotion, onboarded a senior hire, and introduced sprint working agreements and a complexity-weighted help desk rotation.

• Led a team of 8 engineers handling both project work and interrupt-driven operations for enterprise search deployments across multiple clouds.
• Owned development, staging and production environments for enterprise clients on Kubernetes, using Terraform, Argo CD, Helm and custom Python and Go tooling to meet a 99.99% SLA.
• Introduced a structured incident response and root-cause-analysis process, with monitoring and alerting in Prometheus, Grafana and PagerDuty.
• Directed the design of an internal toolbox that consolidated a dozen operational tools into one microservices web application, adopted across teams.
• Led the high-level design of Kubernetes environments, including backup and disaster recovery for stateful workloads, and built reusable Terraform modules and Helm charts adopted as organizational standards.
• Worked with senior leadership to align the operations roadmap with company goals.

• Designed highly available AWS infrastructure (EC2, EKS, ECS, RDS, S3) for government healthcare applications with strict compliance requirements.
• Ran production Kubernetes on EKS with both Fargate and managed node groups.
• Built immutable dev, stage and prod environments with Terraform workspaces and reusable modules, and deployed redundant Azure infrastructure (AD, Key Vault, Storage, VMs, Functions) with Terraform.
• Containerized React, Drupal and Java applications with hardened multi-stage Docker builds and security scanning in CI.
• Automated JMeter load testing in Jenkins so every release was validated against expected concurrency before going live.

• Built and ran AWS infrastructure (EKS, EC2, RDS, S3, API Gateway, ALB) for high-traffic e-commerce and logistics clients, including containerizing a legacy .NET 3.5 application onto EKS.
• Created reusable CloudFormation templates and Ansible playbooks for single-command deployment of load-balanced Java stacks with Jenkins CI/CD.
• Deployed Kubernetes with KOPS and Terraform, with Spinnaker for canary and blue-green releases, and built pipelines for PHP, Python, Go and Node.js services on GitLab, GitHub, Jenkins and CircleCI.

• Built a private cloud on a VMware cluster behind redundant Cisco ASA firewalls, hosting a multi-tenant platform (RemoteApps, Exchange, SharePoint) for 1,000+ users.
• Engineered a ZFS snapshot replication backup product with 30-day retention and minimal bandwidth, and a Splunk health dashboard fed by SNMP and PowerShell.
• Led AD migration, disaster recovery and P2V projects as engineer and project manager.