15.10.2026 aktualisiert
Premiumkunde
100 % verfügbarSenior DevOps / Platform Engineer - Kubernetes, GitOps, IDP, MlOps Cloud-Kostenoptimierung
Berlin, Deutschland
Deutschland +2
Bachelor Engineeringinfo: Deutschland, Österreich, Schweiz
Über mich
Platform engineer: I cut Kubernetes/cloud cost and help teams ship faster. 57% cloud cost cut (PwC), deploys from 1/sprint to 1-4/day (fintech), 50+ clusters migrated to one Terragrunt structure via AI agentic workflows. AWS, Azure, ArgoCD, Backstage, Go. Possible start structure: 2-week assessment
Skills
KeycloakData MeshBackstageMLOpsAzure DevOpsKubernetes-SicherheitBash ShellCloud-EngineeringDevOpsDjangoDNSDataOpsGoogle CloudCloud PlatformIstio
PLATFORM ENGINEERING & INTERNAL DEVELOPER PLATFORM
- Design and rollout of Internal Developer Platforms with Backstage: golden paths, templates, self-service for product teams
- GitOps with ArgoCD and FluxCD, Helm, ArgoWorkflows; deployment frequency raised from 1 per sprint to 1-4 per day at 99% success
- Kubernetes (AWS EKS, Azure AKS, self-managed): multi-cluster operations, zero-downtime migrations, Istio, Cilium
CLOUD COST OPTIMIZATION (FINOPS)
- 57% cloud cost reduction without loss of security, reliability or quality: rightsizing, autoscaling, resource automation, cost allocation
- AWS and Azure architecture, multi-account and multi-subscription setups
INFRASTRUCTURE AS CODE
- Terraform, Terragrunt, OpenTofu, Atlantis; reusable module libraries (VPC, IAM, S3, RDS, KMS, Kubernetes)
- Automated infrastructure testing with Terratest, 85% e2e coverage
- Standardization of Terragrunt repos and modules, migration of 50+ clusters to one target structure, including dependency resolution and safe state migration
AI-ASSISTED INFRASTRUCTURE & AGENTIC WORKFLOWS
- Agentic LLM workflows for large-scale infrastructure migration: code transformation, validation via plan diffs, human-in-the-loop for risky changes
- LLM observability with LangFuse; evals and guardrails for AI-generated changes
- Self-hosted LLM inference (llama.cpp) with failover gateway, MLOps tooling (MLflow, Kubeflow)
DATA PLATFORM & DATA MESH
- Self-service data mesh on Kubernetes: Spark, Trino, Hive, Ranger, Superset, Kafka, MinIO, Jupyter
- Data product time-to-market reduced from days to hours
- Speaker at FWDays Architecture 2023: "Data Mesh in Kubernetes
OBSERVABILITY & RELIABILITY (SRE)
- Prometheus, Grafana, Loki, Alertmanager, ELK; SLO-based alerting
- Incident detection and recovery time (MTTD/MTTR) reduced by 50%
CI/CD & SECURITY
- GitLab CI, Jenkins, release and dependency automation
- Keycloak, Vault, least-privilege IAM, secrets management
SOFTWARE ENGINEERING
- Go (internal tooling: release manager, DNS automation, self-service portal), Python, TypeScript, Bash
- Background in backend development: microservices, Kafka, PostgreSQL, MongoDB, WebRTC
LEADERSHIP
- Grew a platform team from 4 to 10 engineers, multi-year roadmap delivered on time
- Mentoring, internal Kubernetes workshops, stakeholder alignment, shift from consultants to permanent hires
LANGUAGES
English - fluent (preferred project language), German - B1, Ukrainian - native
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2022
Sprachen
DeutschgutEnglischMutterspracheUkrainischMuttersprache
Projekthistorie
- Delivered and continuously refined self service Data Mesh solution utilising Open Source components and Cloud provider capabilities
- Improved time to market, reliability, usability of data products by introduction of self service Internal Development Platform
- Automated data products creation and maintenance by templating, IDP integration and continuous refining
- Reduced cloud bill by 57% without compromising on reliability or products quality
- Automated and improved release process(weekly -> daily), dependency management
- Optimised onboarding, upskilled the team, built relationships with surrounding teams, stakeholders