Sriram Rangaswamy

Led a 5-member engineering team owning the end-to-end observability deployment strategy for Kubernetes workloads on Azure and GCP (AKS/GKE). Architected a unified multi-cluster observability platform using Prometheus, Grafana Alloy, Vector, cAdvisor, and OpenTelemetry; improved service visibility and reduced MTTR by ~40% while maintaining ~99.5% uptime. Delivered enterprise IaC with Terraform across Azure and GCP using reusable versioned modules and environment isolation (dev/perf/prod). Implemented ML platform enablement on Vertex AI (Vertex AI Pipelines/Kubeflow) for training, evaluation, model versioning, and scalable batch/online inference. Strengthened security and access with Entra ID/IAM, Vault/secret manager patterns, RBAC, and automated TLS/credential rotation. Enabled safe rollouts using Istio traffic strategies (canary, blue/green, retries, circuit breaking) and integrated DevSecOps scanning and policy enforcement into CI/CD. Implemented self-monitoring and dashboarding in G

Designed parameterized CI/CD pipelines for Java/Python services (Jenkins/GitHub Actions) using automated tests and quality gates to reduce build/deploy cycle time by ~20% and release defects by ~30%. Implemented promotion pipelines across Dev/QA/Prod with approvals and rollbacks to reduce change failure rate by ~10% and improve audit readiness through structured logging. Containerized microservices with standardized configurations to reduce environment drift. Designed and maintained CI/CD-integrated unit/integration testing frameworks (JUnit, SonarQube, Karate, Pact) to improve release stability and deployment confidence.

Led monolith-to-microservices modernization on AWS/Azure using Kubernetes and ArgoCD; enabled blue/green and canary deployments, reducing downtime per release by ~40%. Automated infrastructure using Terraform/ARM with HA/DR for EKS/AKS, cutting provisioning time by ~30% and improving reliability (RTO/RPO) by ~90%. Built ETL/data stack with Glue, Athena, Redshift, DMS, and Cosmos DB using serverless workflows (Lambda/Python) to reduce data processing costs by ~50% and improve time-to-insight by ~20%. Designed Aurora RDS environments and managed data migration using DMS. Optimized queries and data models for performance and cost. Improved uptime from 99.5% to 99.95%, reduced MTTR by ~40% through monitoring/alerting and automation, and eliminated ~70% of manual operational toil using self-service tools.