50% EKS Cost Reduction – Migrating from Cluster Autoscaler to Karpenter
Migrated a production EKS cluster from Cluster Autoscaler to Karpenter, aggressive bin-packing and Spot instances to slash compute costs by 50%.
Migrated a production EKS cluster from Cluster Autoscaler to Karpenter, aggressive bin-packing and Spot instances to slash compute costs by 50%.
Architected a centralized secret management strategy using AWS Secrets Manager and Parameter Store, removing all hardcoded credentials from codebases and build artifacts.
Migrating from a single-node Jenkins bottleneck to scalable, ephemeral GitHub Actions runners on EKS Auto Mode. Reduced queue times to zero and cut costs by 70%.
End-to-end migration of legacy monolithic applications from EC2 to a containerized microservices architecture on Amazon EKS. Achieved zero-downtime deployments and a 30% reduction in infrastructure overhead.
Migrated backend evaluation workflows from Apache Airflow to AWS Step Functions, eliminating concurrency limits and reducing costs by 60%.
Decoupling data protection from instance performance with a distributed, timezone-aware Step Functions workflow for 50+ tenant databases.
Cost-optimized data orchestration using AWS Step Functions, Lambda, and EventBridge. Replaced compute-heavy ETL jobs achieving 70% cost reduction and 40% faster execution.
I take on a limited number of strategic implementations per quarter. If you're looking to optimize your cloud infrastructure, reduce costs, or streamline DevOps processes, let's connect.