Daniel Raș

About

I’m Daniel, a platform / DevOps engineer from Chișinău, Moldova. I’ve spent about seven years in software. For the last four-plus I’ve owned cloud, Kubernetes and enterprise deployment end to end.

I started as an application developer: front end, then iOS, then .NET backend. I moved to the platform side because I like to understand how systems actually work, not just operate them by instinct.

What I do

In my current role I lead platform and infrastructure for a large microservice product. It runs on-prem, airgapped, and in private clouds (AWS, Azure, OpenShift) for large enterprise customers. I wrote the umbrella Helm chart that runs in customer production. When the product wasn’t ready for OpenShift, I rebuilt the container images to meet its security policies so the deployment could go live.

My pattern across roles is research → proof of concept → adopted standard. That’s how the company got:

  • its observability stack (OpenTelemetry with Grafana, Loki, Tempo and Prometheus, and later a lightweight VictoriaMetrics setup),
  • its internal developer portal (Backstage),
  • its API gateway strategy and integration-testing approach.

I also do the less glamorous work that keeps enterprise deployments alive. I ended a recurring production outage by moving telemetry storage to S3, ran an end-to-end customer data recovery, and replaced a commercial gateway operator with an in-house one to remove a licensing risk.

Strongest at: Kubernetes across distributions (EKS, AKS, OpenShift, RKE2, K3s), infrastructure as code (Terraform, Ansible, Helm), GitOps (ArgoCD), observability, and production reliability and incident response.

Where I’m heading

I’m pointing this platform depth at AI. I’ve built and deployed an internal RAG system over our codebase that colleagues now use. It maps repos to components, generates component knowledge with an LLM, and provides search, a chat interface and an MCP server. Next is the infrastructure layer: self-hosted model serving and GPU orchestration on Kubernetes.

This blog is where I write about that: AI agents, platform engineering, and the setups I build along the way.

Elsewhere