At Quess Singapore, I work on production infrastructure and GPU/AI platform operations, operating GPU fleets on Kubernetes and Slurm. I maintain the NVIDIA GPU stack, including CUDA upgrades, GPU-aware scheduling, and MIG partitioning.
I support distributed training and inference on Kubeflow, Ray, KServe, and vLLM, and deliver applications through GitOps with Argo CD. I also build observability with tools including Prometheus, Grafana, and Datadog.
Earlier, I worked in cloud operations and DevOps engineering at Accenture, and in production support and Unix/Linux systems administration at Coforge, Capgemini, and OIT. My work has included infrastructure automation, incident response, and administering production server estates.
