Dmitrii S.

I’ve spent years helping teams move from infrastructure that grew by accident to deploys nobody’s afraid of, environments that match, and a cloud bill you can explain line by line. I give it a real architecture that is documented and reproducible, then hand it over to your team. One CI/CD pipeline I built became the standard for every repository the client had, and new repos now start from it. When a Kubernetes upgrade broke a client’s ingress and SSL certificates, I brought production back. I’ve also built access control with Terraform and Auth0. As one client wrote: “I have never had an overseas contractor be so focused on communication.” Health-tech is where most of my recent work sits. I run infrastructure for a remote patient monitoring and virtual care platform with customers in Israel, Europe, the UK and the US. Its clinical layer is HAPI FHIR on FHIR R4, holding 14M+ Observation records from connected medical devices, so a slow query or a failed deploy ends up in front of a clinician. Workloads run on Kubernetes with Helm and Flux (GitOps). Releases go through GitHub Actions with SAST, container image scanning, SBOM generation and SonarQube quality gates, across separate Development, Staging, UAT and Production environments. The project operates under ISO 27001 and ISO 13485, so every production change needs traceability, change control and a documented release. I build pipelines that produce that evidence on every deploy, so the team isn’t assembling it the week before an audit. On another healthcare platform I rebuilt a legacy AWS environment with Terraform and FluxCD, fixed the image tagging that was breaking deliveries, automated DNS and certificate issuance, and stood up a dedicated environment for a hospital deployment. Most of the infrastructure I take over runs on AWS (EC2 or ECS, RDS, S3). When you’re already on Azure or Google Cloud, I work the same way there and match your cloud instead of pushing my own. The pipeline uses Terraform for the infrastructure and GitHub Actions or GitLab CI for the deploys, with Docker or Kubernetes wherever the workload needs it. I turn hand-built servers into code with Terraform and Packer, so they can be rebuilt from code instead of from memory. We start with an infrastructure audit that shows what’s actually broken. After that, I’ll walk you through the on-call model and SLA for 24/7 DevOps support. When the work is done, I hand over ownership explicitly, not just access.