At Vela Software, I architect and operate Kubernetes environments across AWS, GCP, and bare-metal GPU nodes, sustaining 99.9% uptime for production platform services. I’ve configured AMD Instinct and NVIDIA GPU clusters with ROCm and vLLM, resolving multi-GPU scheduling bottlenecks to improve LLM inference throughput.
I’ve also deployed GitOps delivery pipelines with ArgoCD and GitHub Actions, and automated bare-metal provisioning with MAAS and Terraform. Earlier, at Tkxel and Avanceon, I worked on high-availability infrastructure, security monitoring, Kubernetes, and observability.
