Sai Ram

Independent Cloud Platform Consultant, Jan 2022 - Present, India (Remote) MLOps / AI Infrastructure - GPU Inference Platform, 2024 - Present • Designed and built an EKS-based GPU inference platform using vLLM, Karpenter, Terraform, and ArgoCD for GitOps delivery. • Shifted eligible inference workloads from on-demand GPU instances to Spot-backed capacity with graceful interruption handling, reducing inference compute cost by ~40%. • Added autoscaling, model-loading optimization & Prometheus-based monitoring for serving workloads. • Built self-service ML training and deployment pipelines using Kubeflow, ArgoCD, and GitOps; reduced typical model delivery time from days to hours for the data science team. FinTech / Payments - Platform Reliability & Compliance, 2023 - 2024 • Designed multi-region disaster recovery and compliance automation for payment workloads on AWS and Kubernetes, supporting documented availability targets and audit requirements. • Implemented Kubernetes RBAC, secrets management via Vault, vulnerability scanning, and CI/CD compliance gates aligned with PCI-DSS controls. • Produced runbooks, alerting, and operating-model documentation for production handover to the client team. FinOps / Multi-Cloud Cost Optimization - AWS to OCI, 2023 • Designed AWS-to-OCI architecture using FastConnect; kept latency-sensitive services on AWS and shifted batch and storage-heavy workloads to OCI. • Reduced combined compute and egress spend by ~30% through workload placement, rightsizing, and network design; documented migration plan and rollback process. Enterprise - Terraform Modernization, 2022 • Refactored legacy infrastructure into a modular Terraform library with environment standards, autoscaling, and quota controls; reduced unplanned cloud spend and shortened environ