Work

Since April 2026 I combine SRE with building the company’s own AI platform: an LLM inference API served from our own GPU fleet.

• Run the LLM inference platform in production on Kubernetes (vLLM, LiteLLM): model sizing by VRAM and KV-cache, tuning with real metrics and evaluation of new models.
• Build the platform in Go and SvelteKit/Astro: RAG chat, Stripe billing, a referral system, feature flags and a TUI.
• Design the platform’s observability and alerting with VictoriaMetrics, Grafana and VictoriaLogs.
• Own the reliability of production PostgreSQL databases: backups, recovery and performance.
• In parallel, cloud infrastructure for clients on GCP and AWS.

• Manage Kubernetes clusters across AWS, Azure and DigitalOcean for client and internal environments, ensuring high availability and operational reliability.
• Deploy Kubernetes clusters and CI/CD pipelines from scratch for clients, with Terraform and Scalr as the main IaC tooling.
• Migrate production workloads from EC2 instances to Kubernetes, designing custom Helm charts tailored to each project.
• Standardize CI/CD pipelines on GitLab CI, GitHub Actions and Bitbucket Pipelines.
• Implement secret management with HashiCorp Vault integrated with Kubernetes through the Vault Secrets Operator.
• Deploy and maintain applications with Helm and ArgoCD following GitOps principles.

• Managed 30+ Kubernetes clusters across AWS, GCP, Azure, and on- premise (k3s/kubeadm), ensuring high availability and disaster recovery for client environments.
• Operational Architecture: Collaborated with cross-functional teams to design scalable, secure architectures aligned with business requirements.
• Tooling & Automation: Focused on adoption of OpenTelemetry, FluentBit, Loki, and Tempo to centralize logging/tracing, reducing incident resolution time.
• CI/CD: Standardized secure pipeline designs using Gitlab, ArgoCD, Jenkins and Github Actions enforcing best practices (Trivy for vulnerability scanning, SonarQube for code quality).
• Observability: Built and maintained monitoring stacks (Prometheus/Thanos, Grafana, ELK).
• Configured and implemented Kong for API gateway functionality and load balancing, and Keycloak for robust identity management,
• Mentorship: Provided technical guidance to junior engineers on Kubernetes, Helm, and GitOps workflows.

• Negotiate with vendors regarding software services and products.
• Implement the necessary solutions to meet the needs of the company and its users.
• Manage the IT team and coordinate the work of the area.
• Manage GitLab CI and ArgoCD deployments.
• Migrate on-premise services with Docker to a cloud infrastructure with Kubernetes.

• Definition and improvement of CI / CD processes (Gitlab, Drone, Gitea )
• Documentation management (WikiJS)
• Monitoring by Prometheus + Grafana
• Log management through Promtail + Grafana Loki + Grafana
• Container management through Docker with Docker-compose, Traefik as reverse proxy and loadbalancer.
• Tutoring of students in computer engineering internships.

After-sales technician:

• Incident management.
• Warranty management.
• Customer service.
• Provide comprehensive support and problem resolution for customers.
• Stock management.