At Kaspi Bank, I operate production Kubernetes clusters supporting high-throughput AI/ML workloads, and manage NVIDIA GPU infrastructure across B200, H200, and L40S accelerators.
I migrated a high-traffic computer vision inference workload to a newer GPU cluster, resolving Triton compatibility and model warm-up issues while maintaining production service continuity. I also fixed a Kubernetes networking scalability bottleneck in the Cilium layer.
Previously, at Netcracker Technology, I developed reusable CI/CD workflows and automated customer environment provisioning and deployments with Terraform, Ansible, Kubernetes, and ArgoCD. I’m focused on production reliability, incident response, and automation for platform and AI infrastructure.
