7+Years specializing in cloud and DevOpsAWS · Azure · GCP
5–7Engineers led across concurrent client accountsThree direct reports
5+Client platforms architected from scratchGreenfield
3Clouds run in production, not just evaluatedMulti-cloud
ZeroDowntime incidents through Black Friday peak trafficSpend held under cap
Selected work
5 systems
Higher-education AI platform
Folio3 / client platform
Onboarding a new university is a GitHub Actions run, not a manual build-out.
ComplianceFERPA, HECVAT-assessed
Platform
Multi-tenant AI SaaS, Y Combinator-backed
Role
Infrastructure owner across dev, staging and prod
Stack
ECS Fargate
Aurora PostgreSQL
ElastiCache Redis
Terraform
AWS CDK
Redesigned the platform into per-school ECS stacks backed by shared Lambda data-sync services for Salesforce and SIS/LMS — stronger tenant isolation, less duplicated shared-service overhead.
Drove donor and account data-sync with AWS Step Functions using Map and Parallel states, fanning Lambda invocations across large datasets with controlled retries and concurrency limits.
2 moreCollapse
Held sole ownership of infrastructure decision-making and wrote the disaster-recovery, incident-response and operational runbooks, with no interruption to service.
Migrated ECS Fargate workloads to arm64/Graviton across customer stacks with no disruption, and stood up PostHog and Langfuse for product and LLM observability.
Healthcare care-coordination platform
Folio3 / client platform
Standing up a new environment is a module call. Five of them now run behind a single load balancer.
ScaleOne ALB fronting five application stacks
Role
Sole DevOps engineer on the account
Stack
EKS
Terraform
Helm
InterSystems IRIS
Elasticsearch
GCP
Packaged the EKS module with its core add-ons — CoreDNS, VPC CNI, EBS CSI and metrics-server — plus submodules for Cluster Autoscaler, AWS Load Balancer Controller and ExternalDNS, so scaling and DNS come wired in with the cluster.
Re-platformed a production InterSystems IRIS database cluster from ECS to EKS, replacing hand-maintained YAML with Terraform-managed Helm releases and consolidating legacy ECS, ALB and VPC resources into one region.
Shipped Snowstorm, a SNOMED CT terminology server with FHIR APIs, onto EKS via a custom Helm chart backed by Elasticsearch — working through JVM, startup-timing and storage-sizing issues to production readiness.
4 moreCollapse
Built the GCP footprint from scratch across dev, preprod and prod (VPC, subnets, NAT gateway, flow logging) plus a reusable Terraform SFTP-as-a-service module with per-environment service accounts and versioned GCS buckets for secure patient-data imports.
Built developer SSH onboarding on short-lived OIDC credentials issued by GitHub Actions, provisioning EC2 users through SSM RunShellScript with failure detection and duplicate-user guards.
Reconstructed a legacy Rails application's Helm chart by working backwards from its CI/CD configuration.
Tuned observability and access control: ECS Container Insights, least-privilege IAM/SSM task roles, and right-sized EC2 and Auto Scaling groups.
On-prem AD to Entra ID
Folio3 / migration lead
Retired a company's on-premises Active Directory for good, sequencing the cutover so business-critical systems stayed up throughout.
Plan15+ steps across four workstreams
Scope
Tenant-to-tenant Azure migration retiring on-premises Active Directory
Role
Technical lead and primary client contact
Stack
Microsoft Entra ID
Azure Virtual Desktop
FSLogix
Intune Autopilot
Directed the Azure Virtual Desktop build-out with FSLogix profile management and load-based session-host sizing, and planned Intune Autopilot enrollment to replace AD-joined devices with a cloud-managed fleet.
Oversaw migration of the core business services — application servers, databases, mail and SharePoint — coordinating each across planned cutover windows so business-critical systems stayed available.
1 moreCollapse
Ran client communications throughout, turning technical milestones into status updates and managing risk and timeline expectations across a multi-week migration.
Legacy monolith re-architecture
Folio3 / client platform
Every build used to drift from the last one. Containerizing the monolith made development, staging and production identical.
ShapeSingle host to isolated microservices
Platform
Ran directly on a bare RedHat 8 host
Role
Re-architecture lead
Stack
Docker
Nginx
Python
Mono
MongoDB
Replaced manual yum installs, from-source Python compilation and hand-configured Apache modules (mod_ssl, mod_wsgi, mod_mono) with reproducible Docker images, ending environment drift between builds.
Fronted the application modules with Nginx routing through a central API to horizontally scaled calculation-engine containers, so compute-heavy work scales independently of the web and API layers.
1 moreCollapse
Cut the attack surface of the original single-host deployment by enforcing container-level isolation and restricting resource access to essential components only.
Deployment Dashboard
Personal project / solo build
Replaces the Slack thread where deploys used to be tracked. Shows what's live where — and which commits reached staging but never production.
WhatDeployment tracking for GitHub organizations
Role
Full-stack and infrastructure, start to finish
Stack
Next.js 14
TypeScript
PostgreSQL/Prisma
Docker Compose
AWS CDK
Syncs continuously against the GitHub API through Octokit — supporting both PAT and GitHub App authentication — with background jobs, so status stays current without manual polling.
Runs on PostgreSQL and Prisma, containerized with Docker Compose behind a Makefile-driven workflow, with integration and unit tests covering its own AWS CDK stack and GitHub Actions shipping both the app and its infrastructure.
Also delivered
Trading-card valuation platform
Designed the whole cloud footprint in Terraform for an AI trading-card valuation platform — EKS, ECS on both EC2 and Fargate, Amazon MQ and ElastiCache — GPU-optimized around a production Milvus vector database.
Multi-site Magento
Re-platformed a monolithic Magento install (2.4.1 to 2.4.5-p1) onto containerized ECS Fargate with centralized CloudWatch dashboards and a Bitbucket to CodeBuild to CodePipeline flow, holding combined daily AWS spend under an agreed cap.
Test-prep platform
Built AWS from nothing for a PHP/Laravel platform — VPC, ECS Fargate, ALB, RDS and CloudFront — behind a CodePipeline-driven flow that deploys Docker images automatically.
K-12 learning platform
Moved a Moodle application off EC2 onto ECS Fargate, containerizing it and provisioning RDS, ElastiCache and CloudFront behind an automated Bitbucket pipeline.
Experience
Folio3 Software · 2015 — now
Jul 2022 — present
Lead DevOps Engineer
Lead and mentor five to seven engineers, three of them direct reports, across concurrent client accounts while personally owning production infrastructure for two SaaS platforms. Built a GPU-backed ECS/EKS compute platform using custom capacity providers, EC2 warm pools and RabbitMQ queue-depth auto-scaling via Lambda, and stood up an in-house Kubernetes cluster with Nginx Ingress HTTPS routing that internal teams now self-serve. Directed peak-traffic readiness for e-commerce clients through Black Friday, aligned access control, secrets management and backup policy to ISMS and GDPR, and run day-to-day Azure administration (D365 environments, VMs, MS SQL, Redis) plus Office 365 and Partner Portal access for several client accounts.
Jan 2021 — Jun 2022
Senior Systems Engineer
Designed the network infrastructure for a new 13-floor building, including a local datacenter housing 20+ rack servers. Rebuilt a healthcare platform's AWS footprint from a single EC2 instance into an auto-scaling, HIPAA-aligned architecture by scripting AMI creation and launch-template updates, and moved a WordPress workload from an Azure VM to App Service with Azure CI/CD pipelines. Managed Azure sponsorship subscriptions and a BitLocker/Group Policy rollout, and evaluated and deployed monitoring and endpoint tooling — Zabbix, PRTG, Uptime Robot, Kaspersky EDR.
Jan 2018 — Dec 2020
Systems Engineer
Administered Linux and Windows environments across LAMP/LEMP stacks and migrated on-prem physical and virtual machines to AWS and Azure with Veeam replication. Deployed off-site Veeam disaster recovery, a self-hosted GitLab server and a site-to-site VPN across two offices. Replaced Xen virtualization with Hyper-V and ran the dev/QA VM estate, and remediated several ransomware incidents by isolating affected systems and rebuilding from clean backups, with minimal data loss.
Dec 2015 — Dec 2017
System Support Engineer
End-user hardware and software support. Deployed WSUS and Windows Deployment Services to automate OS installation and patch compliance, and migrated Windows Server 2008 to 2012.