• 🛠️ Senior DevOps Engineer based in Tehran, Iran — 8+ years across cloud, bare-metal, and platform engineering.
• ☸️ Operate four production Kubernetes clusters across three regions (Hetzner RKE1/RKE2, EKS) via Rancher.
• 🤖 Build LLM & MLOps infrastructure — Llama 3 / Mistral on Kubernetes with vLLM, GPU scheduling, internal AI-agent developer platform (Coder v2).
• 🔐 DevSecOps — bastion design, two-tier SSH, OWASP CRS, hardened Ubuntu baselines.
• 🧰 Author of kubeagent — a Kubernetes troubleshooting agent in Go.
• 🌐 Landing page: k8sproject.top

Read-only Kubernetes troubleshooting agent (Go / client-go) that scans clusters and explains pod failures — CrashLoopBackOff, ImagePullBackOff, OOMKilled, Pending/Unschedulable — with prioritized P1-infra → P2-workload reports.

• 🔎 Platform detection (CNI, ingress, storage, runtime, cloud) + live CPU/memory via metrics-server
• 🩹 Guarded, reversible --fix remediation (rollout undo, uncordon) applied only after confirmation
• 🧠 Optional --explain AI mode — privacy-safe findings to the Claude API (never raw specs/secrets)
• 🔒 --lint-secrets and NetworkPolicy hints; chaos-testing suite; v1 + v2 shipped

Cloud & Orchestration

IaC, CI/CD & Config

Observability & Data

Languages & OS