Senior DevOps / Site Reliability / Platform Engineer with 10+ years of experience, based in Turkey and working remotely.
I build and run cloud platforms for high-traffic fintech and iGaming products: multi-cluster Amazon EKS with disaster recovery, GitOps with Argo CD, Terraform / Terragrunt for everything, and observability with Prometheus, Grafana, Loki and VictoriaMetrics. Lately I've been adding LLM-assisted tooling to incident response.
| Project | What it shows |
|---|---|
| terraform-aws-eks-platform | Production-style EKS platform: VPC, EKS, Karpenter, platform add-ons and Argo CD bootstrap; Terragrunt live layout for dev / prod / prod-dr |
| terraform-azure-aks-platform | Self-service AKS platform: Entra ID + namespace-scoped Azure RBAC, workload identity, Node Auto-Provisioning, Flux via the AKS extension, Istio Gateway API, Gatekeeper policies tested with gator, one-file team onboarding |
| terraform-azure-landing-zone | Azure landing zone: CAF management groups, policy-as-code by management group, hub with Azure Firewall + DNS Private Resolver, subscription vending with azapi from one YAML file per team |
| terraform-aws-eks-ai-cluster | GPU / LLM inference on EKS: Karpenter GPU NodePools, NVIDIA device plugin + DCGM alerts, S3 model store via Mountpoint, vLLM scaled by KEDA on request queue depth |
| k8s-postgres-platform | PostgreSQL on Kubernetes with CloudNativePG: three zones, quorum commit, PgBouncer, PITR, major upgrades and a Helm chart for tenant databases; failover (~16 s, zero lost writes) measured in CI |
| rds-to-k8s-migration | RDS PostgreSQL -> CloudNativePG under live writes: logical replication, cutover with ~2 s of frozen writes and checksum checks, a rollback through reverse replication that loses nothing, and a point-in-time restore drill; RDS MySQL -> MOCO through a GTID relay in progress |
| terraform-cloudflare-edge | Cloudflare edge as code: WAF + OWASP, geo-fencing with partner bypass, rate limits, Logpush to tiered S3 and OpenSearch, Tunnel + Access for internal tools and a self-hosted LLM, AI Gateway |
| cloudflare-edge-platform | Cloudflare in front of Kubernetes with no public load balancer: outbound-only Tunnel with kind e2e checks, WAF and rate-limit rules as reviewed YAML with 29 Terraform tests, Logpush to R2 queried with DuckDB, incident runbooks |
| edge-llm-gateway | OpenAI-compatible LLM gateway on Cloudflare Workers: per-team keys, rate limits and monthly token budgets in Durable Objects, fallback with circuit breakers, streaming usage metering, PII redaction; 62 tests in the Workers runtime, Terraform for the edge |
| eks-to-onprem-repatriation | Phased EKS -> on-prem migration: Talos, Cilium BGP, Rook-Ceph, CloudNativePG logical replication with a rollback path, Strimzi MirrorMaker 2, Cloudflare weighted traffic shift, TCO model |
| gitops-multicluster-dr | Argo CD ApplicationSet matrix generators across environments and clusters, plus drswitch, a Go CLI for single-switch DR failover |
| aiops-incident-assistant | Alertmanager webhook → logs and metrics context → LLM incident summary in Slack, with secret redaction |
| postgres-ops-agent | Read-only MCP server and Claude Code skill for Postgres incident triage: least-privilege identities, human-approved actions with an audit log, fault-injection evals |
| platform-ci-templates | Reusable GitHub Actions, GitLab CI, Bitbucket Pipelines and Jenkins templates: OIDC to AWS, SBOM, image signing, GitOps promotion |
| k8s-observability-stack | GitOps-ready Prometheus, Grafana, Loki and VictoriaMetrics with SLO burn-rate alerts, rule unit tests and dashboards |
| tf-modules | Small, readable Terraform modules: VPC, EKS, add-ons with IRSA, Argo CD and RDS with External Secrets |
Cloud: AWS (EKS, EC2, VPC, IAM, MSK, RDS), Azure (AKS, Azure DevOps, Azure Policy, Azure Firewall), Cloudflare Kubernetes: EKS, AKS, Helm, Karpenter, KEDA, Cilium, Argo CD IaC: Terraform, Terragrunt, OpenTofu CI/CD: GitHub Actions, Bitbucket Pipelines, GitLab CI, Jenkins, TeamCity, Octopus Deploy Observability: Prometheus, Grafana, Loki, VictoriaMetrics, Elasticsearch, PagerDuty Data and messaging: Kafka, RabbitMQ, MongoDB, PostgreSQL (CloudNativePG), SQL Server Languages: Python, Go, Bash
Certifications: HashiCorp Terraform Associate · CKAD · AWS Cloud Practitioner