• Kubernetes platform engineering (EKS-first) ● Design, build, and operate production-grade Kubernetes clusters (multi-nodegroup, autoscaling, upgrades, cluster add-ons). • Implement intelligent autoscaling using real metrics (queue depth, consumer lag, service latency) via tools like KEDA/Karpenter. • Own AWS environments end-to-end (VPC, IAM, EKS/ECS/EC2, ALB/ELB, S3, Route53, CloudWatch, RDS, SQS, Lambda). • Build reproducible infrastructure using Terraform, with strong review + change management practices. • Implement backup/DR patterns (e.g., snapshots, retention, automation) and safe rollouts. • Design infrastructure for data-intensive workloads: high-throughput ingestion, batch processing, and real-time streaming. • Understand and operate distributed systems at scale — consensus, partitioning, replication, and failure modes. • Build and maintain infrastructure for data pipelines, vector databases. • Design for horizontal scalability, ensuring systems handle growing data volumes and user traffic gracefully. • Build/own monitoring + logging from scratch and make it actionable (Prometheus/Grafana, ELK/EFK, alerting). • Define/partner on SLI/SLOs and incident response practices; improve reliability with data-driven changes. • Establish performance testing and production-like load testing environments. • Continuously reduce AWS spend via right-sizing, Spot strategies, reserved capacity planning, and architecture improvements. • Partner with engineering teams to diagnose bottlenecks (db queries, caching, queueing) and propose scalable solutions. • Optimize infrastructure costs for data-heavy workloads (storage tiering, compute scheduling, GPU utilization). • Improve cloud and cluster security posture (IAM, network policies, secrets management, least privilege). • Support SOC2 readiness/execution (controls, evidence automation, operational hardening). • Implement access management patterns.