Platform Engineer @ project44 · Chicago, IL

Dmytro Dmytrush

DevOps / Platform Engineer — AWS · Kubernetes · Terraform · Agentic AI

5+ years building cloud infrastructure, Kubernetes platforms and CI/CD automation on AWS. AI-native engineer — 20+ custom agent skills across Claude Code and Codex for infrastructure and deployment work, using MCP tools and AI-assisted review workflows to accelerate delivery and reduce operational toil.

See my work Say hello
CKA CKAD AWS Solutions Architect – Associate HashiCorp Terraform Associate
99.95%availability 14×traffic bursts absorbed 12EKS clusters 50+apps supported
By the numbers

Impact, quantified

0%
service availability
0×
peak traffic absorbed
0%
fewer on-call pages
0+
apps supported
0%
compute cost cut
The stack

What I work with

Five-plus years of picking tools that make platforms boring in the best possible way — predictable, observable, and safe to change.

Cloud & Containers

Multi-account AWS estates, EKS clusters with mixed on-demand and Spot capacity, and least-privilege access via IRSA.

AWS EKSIAM / IRSACloudWatch Secrets ManagerS3RDS ALBKubernetesDockerLinux Rancher

IaC & GitOps

Everything reproducible, reviewed, and shipped through a pull request.

TerraformHelmArgo CDFlux CD GitHub ActionsGateway APIAWS LB Controller

Observability & Reliability

SLOs and error budgets instead of noisy thresholds — alerts that mean something at 3am.

PrometheusGrafanaAlertmanager OpenTelemetryJaegerElasticsearch KibanaFluentd SLOs / error budgetsIncident responseOn-call

AI & Automation

Agentic tooling used daily on real infra work — runbook generation, alert summarization, root-cause analysis.

Claude CodeMCPCustom skills AWS BedrockContext engineering BashPython

Service Mesh & Traffic

mTLS between services, progressive rollouts, and ingress that stays declarative instead of hand-tuned.

IstioIstio GatewayKialiGateway API

Security & Policy

Guardrails enforced in the cluster instead of in a wiki — admission policy, encrypted secrets in Git, identity-based access.

KyvernoSealed SecretsZero TrustIRSA

Scaling & Resilience

Capacity that follows demand instead of guesswork, plus backups that have actually been restored.

KarpenterKEDAHPAVelero
Experience

Where I've done it

High-traffic consumer marketplaces and supply-chain systems — the common thread is turning manual work into platform.

Platform Engineer · project44

Sep 2023 – Present

Chicago, IL

  • Operated platform infrastructure for 50+ containerized applications across 12 EKS clusters and multiple AWS accounts, serving 9 engineering teams (~100 engineers).
  • Built reusable platform components — Terraform modules, Helm charts, Argo CD ApplicationSets, GitHub Actions workflows — standardizing provisioning and deployments across engineering teams, reducing environment setup from 1–2 days to under 30 minutes while decreasing platform support requests 40%.
  • Implemented observability standards by building SLO dashboards, codifying Prometheus alerts, and maintaining Grafana dashboards, improving production monitoring and incident response.
  • Built reusable AI-assisted DevOps workflows using Claude Code, Codex and MCP integrations to automate runbook generation, summarize production alerts and assist incident triage — accelerating Terraform, Kubernetes and CI/CD work while reducing incident investigation time 30%.
  • Migrated platform networking from ingress-nginx to Kubernetes Gateway API backed by AWS Load Balancer Controller, standardizing traffic routing and separating platform-managed Gateways from application-managed HTTPRoutes.

DevOps Engineer · SeatGeek

Jul 2021 – Sep 2023

New York, NY (Remote)

  • Optimized Terraform-managed Amazon EKS infrastructure supporting production Kubernetes workloads, using On-Demand and Spot node groups to balance application reliability and infrastructure cost.
  • Migrated production AWS infrastructure from CloudFormation to Terraform with zero downtime, improving infrastructure consistency, scalability and operational efficiency.
  • Built autoscaling for the inventory service based on live request volume, not just CPU — scaling 3 → 22 pods in under 2 minutes during 14× traffic spikes on major on-sale days while keeping response times under 400ms.
  • Cut compute costs 35% on quiet days by shifting non-critical workloads to cheaper Spot capacity, while keeping order processing on reliable on-demand instances.
  • Configured and maintained cert-manager, ExternalDNS and Sealed Secrets on Amazon EKS, automating TLS certificate management, DNS updates and secure Kubernetes secret handling.
Deep dives

Selected work, in detail

The full write-ups — architecture, trade-offs, and what actually broke along the way.

SeatGeek · Infrastructure migration

Zero-Downtime CloudFormation → Terraform Migration

Taking a year of accumulated CloudFormation drift into Terraform without downtime — importing VPC, subnets, security groups, IAM roles with IRSA trust policies, the EKS cluster and both managed node groups into state with zero resource replacements. Cross-stack ImportValue dependencies were moved to data sources before any stack was deleted.

TerraformCloudFormationEKS IRSAPrometheus
Read the deep-dive
project44 · Platform engineering

Internal Developer Platform @ project44

Golden-path tooling — Terraform modules, Helm charts, Argo CD ApplicationSets and GitHub Actions templates — serving 50+ applications across 12 EKS clusters and 9 engineering teams. Provisioning dropped from 1–2 days to under 30 minutes, and manual support requests fell 40%.

TerraformHelmArgo CD GitHub ActionsEKS
Read the deep-dive
Dmytro Dmytrush
A little about me

Infrastructure person, people-first

I came into engineering sideways — a finance degree from Lviv, then a hard left into Linux, containers, and eventually the AWS console at 2am. What stuck is the same thing that drew me in: the satisfaction of taking something fragile and manual and making it boring, documented, and self-serve.

These days I spend my time on golden paths — the Terraform module, the Helm chart, the Argo CD ApplicationSet that lets a product team ship without filing a ticket. I'm also deep into AI-assisted operations, building custom Claude Code skills and MCP integrations that shave real time off incident work.

Chicago, IL BS Finance & Credit, Lviv Academy of Commerce 4 active certifications 5+ years in production
Get in touch

Let's talk infrastructure

Open to Platform Engineer and Senior DevOps roles in Chicago or fully remote — and to conversations about platform engineering, SRE, and anything at the intersection of Kubernetes and AI-assisted operations.

contact@dmytro-dmytrush.com LinkedIn