Building fault-tolerant, self-healing infrastructure on AWS and Kubernetes. Fifteen-plus years across cloud, IaC, CI/CD and observability — now building AI platforms: production Kubernetes for GenAI workloads, LLMOps pipelines and agentic tooling.
cloud/k8s → highly-available, multi-account AWS · EKS · service mesh
infra-as-code → Terraform · Ansible · reproducible environments
gitops → ArgoCD · Argo Rollouts (blue-green, canary)
observability → Prometheus · Grafana · Thanos · unified logging + alerting
ai platform → GenAI on EKS · Dify · Onyx · RAG · vector databases
ai agents (now)→ Claude Code · MCP servers · LiteLLM · multi-agent workflows
| Domain | Tooling |
|---|---|
| Cloud & Orchestration | AWS Kubernetes EKS Helm Istio Docker Containerd |
| IaC & Automation | Terraform Ansible Packer CloudFormation |
| CI/CD & GitOps | ArgoCD Argo Rollouts Flux GitHub Actions Jenkins GitLab CI |
| Observability | Prometheus Grafana Thanos ELK Jaeger PagerDuty |
| AI / LLMOps | Claude Code MCP LiteLLM Dify Onyx Karpenter vector DBs |
| Data & Security | PostgreSQL MongoDB Redis Snowflake HashiCorp Vault AWS KMS |
| When | Role | Shipped |
|---|---|---|
| now | Platform Engineer · Studytube | Production EKS platform for GenAI workloads — Karpenter node autoscaling, ArgoCD GitOps, Gateway API, External Secrets, IRSA/KMS security baseline; runs LLM apps (Dify, Onyx) and RAG pipelines end-to-end |
| 2024 → 2025 | Platform Engineer · Bitvavo | HA AWS + Kubernetes for crypto infra; automated 50+ blockchain wallet nodes with Terraform + ArgoCD; progressive delivery via Argo Rollouts; 99.9% uptime with automated recovery |
| 2022 → 2024 | Senior DevOps · DataArt | Multi-account AWS with Control Tower; migrated legacy services to Kubernetes + Istio; unified Prometheus/Grafana/Thanos observability; deploy time −75% |
| 2020 → 2022 | Project Lead & DevOps · Samson-opt / Rework | Led infrastructure automation; established Terraform + Ansible IaC; centralized logging & monitoring |
| 2015 → 2020 | Full-Stack → DevOps Engineer | Moved from application development into infrastructure; built CI/CD across diverse stacks |
Full history on LinkedIn.
- GenAI Platform on EKS — production Kubernetes platform for AI workloads: Karpenter autoscaling, ArgoCD GitOps, Gateway API, External Secrets; hosting Dify, Onyx and RAG pipelines with full IaC (Terraform).
- Agentic Engineering Setup — multi-agent development workflows on Claude Code: custom MCP servers, shared knowledge-graph memory (Graphiti/Cognee), LiteLLM gateway, autonomous CI-verified delivery.
- LLM Infrastructure Automation — automated 50+ wallet-node deployments, integrated MongoDB Atlas with Snowflake, built LLM-specific CI/CD pipelines.
- Kubernetes + Istio Migration — modernized legacy infra onto multi-account AWS with a service mesh and GitOps-driven progressive delivery.
- Unified Observability Platform — centralized Prometheus + Thanos with Grafana dashboards and automated PagerDuty alerting; MTTD −80%.
- Security-First Architecture — Vault-managed secrets, KMS encryption everywhere, self-healing HAProxy with liveness probes.
- DevOps Best Practices — infrastructure patterns & reference automation
- AI Technologies — LLMOps and AI-infra experiments
- LinkedIn — full professional history
$ echo "open to interesting infrastructure & AI-ops challenges"