R Narendran

Architected and managed the high-performance GPU infrastructure supporting the proprietary LLM product (relax.ai), ensuring the scalability and high availability of the multi-tiered conversational AI/ML platform. Orchestrated operators and releases across all active regions, enforcing immutability and governance through ArgoCD-powered GitOps pipelines.

Streamlined scalable multi-tenant AWS EKS platforms using Terraform, embedding security best practices to achieve wiz.io compliance, reducing cloud costs by 60% and audit-ready infrastructure.
Implemented unified observability stack (Prometheus/Grafana), customizing alerting pipelines to reduce Mean Time To Detect (MTTD) critical infrastructure and application anomalies by 40%.

Conceptualized multi-region Kubernetes clusters using Terraform and custom Golang controllers, optimizing production resources by 20%and ensuring high availability.
Implemented a centralized Observability Platform (Prometheus, Grafana, Thanos) for multi-region logs and metrics, reducing critical incident detection time by 50%

Transformed the AI/ML workload environment on AWS EKS by implementing GitOps for deployment automation and enhancing reliability with GPU metric monitoring; slashing storage costs by 73%−78% via strategic migration.

Delivered 10+ multiplatform, full-stack applications, streamlined CI/CD using Docker containerization and Git best practices for deployment on AWS, accelerating delivery and deployment frequency by 63%