10/12/2026 updated
ON
100 % available
Senior DevOps / Platform Engineer | GCP, K8s, GitOps, AI/LLM Infra
Warsaw, Poland
Worldwide
Master of Computer Science, Information Systems - Kherson State Technical UniversityAbout me
Senior DevOps & Platform Engineer with 25+ years in IT. I bridge modern cloud-native architecture (GKE, AWS, GitOps) with robust engineering fundamentals. Proven track record in zero-downtime migrations, AI/LLM infrastructure at scale, API security, and cost optimization.
Senior DevOps / Platform Engineer with 25+ years of IT experience, specializing in production multi-cluster Kubernetes (GKE), GitOps delivery, AI/LLM platform infrastructure, and scalable self-hosted tooling. Experienced in API security, disaster recovery, cloud migration, and vendor lock-in minimization.
Key Technical Expertise:
• Cloud Providers: GCP (GKE, Cloud SQL, GCS, Artifact Registry, Cloud NAT, IAM), AWS (EKS, ECS, Lambda, S3)
• Kubernetes & Cloud Native: Multi-cluster GKE, Linkerd (mTLS, multicluster), Kong Gateway API, ArgoCD, Helm, cert-manager
• AI / LLM Infrastructure: vLLM inference servers (NVIDIA L4/T4 GPUs), Langfuse tracing & evaluation, LLM-powered agents, MCP integrations, n8n
• CI/CD & Security GitHub Actions, ARC self-hosted runners, Vault, Gitleaks, Codecov, Nexus, CORS hardening, mTLS
• Observability & Data: Prometheus, Grafana, ELK Stack (self-hosted), Jaeger, OpenTelemetry, ClickHouse, Apache Superset, DataHub, Redpanda (Kafka)
• IaC & Automation: Terraform, Ansible, Python, Bash, TypeScript/React (tooling UIs)
• Databases & Messaging: PostgreSQL, Redis, RabbitMQ, Temporal, Celery, Elasticsearch
Notable Achievements & Experience:
• Designed and operated 6+ production GKE clusters running Python/Django services and AI/LLM workloads.
• Executed zero-downtime full-stack infrastructure migration from AWS to GCP using Helm, ArgoCD, and Terraform.
• Upgraded API security by migrating traffic to Kong Gateway API with RS256 JWT validation and Cloud Armor edge protection.
• Built and optimized AI/LLM inference platform using vLLM and self-hosted observability tools (Langfuse, ClickHouse).
Experience
Professional experience10+ years
Freelancer since2016
Languages
EnglishFluent
Project history
• Designed, operated, and evolved a production multi-cluster GKE platform (6+ clusters) supporting Python/Django services and AI/LLM workloads on GCP.
• Executed full-stack zero-downtime infrastructure migration from AWS to GCP using Helm, ArgoCD, and Terraform.
• Standardized CI/CD delivery using GitHub Actions, ArgoCD, and Helm; implemented self-hosted ARC runners on GKE.
• Migrated north-south traffic to Kong Gateway API with RS256 JWT validation, HashiCorp Vault secrets management, and Linkerd service mesh mTLS.
• Deployed and operated vLLM inference servers with NVIDIA GPUs (L4/T4) and self-hosted Langfuse for LLM tracing and observability.
• Cost optimization: reduced GCP infrastructure spending by consolidating GCS buckets, tuning metrics cardinality, and setting registry retention policies.
• Architected and built large-scale infrastructure across two data centers in Stockholm using bare-metal Kubernetes, Ceph storage, Proxmox/KVM virtualization, and Ansible.
• Managed self-hosted ELK stack, Grafana/TICK, Icinga2, GitLab CI/runners, Docker, HBase, and Elasticsearch clusters.
• Planned and executed DC-level disaster recovery and failover procedures alongside Linux server and network administration.