Cloud & Infra
AWS multi-account architectures with Organizations, IAM Identity Center (SAML/SCIM via Google Workspace), and cost optimization strategies.
11 years in tech · 6 in DevOps · AWS · EKS · Terraform · Datadog · GitHub Actions
I’m Roger Dávila, a DevOps/Cloud Engineer and Solutions Architect with 11 years in tech and 6 in DevOps. My path started in Java development, moved through data warehousing and solutions architecture, and evolved into DevOps and Platform Engineering. I build secure, resilient, and cost-effective cloud platforms with a pragmatic, business-driven mindset.
Most recently I served as the sole DevOps/Platform Engineer at a RegTech startup, owning the entire platform for multiple development teams — from EKS infrastructure and CI/CD to observability and security. I thrive in environments where autonomy, ownership, and cross-cultural communication are essential.
AWS multi-account architectures with Organizations, IAM Identity Center (SAML/SCIM via Google Workspace), and cost optimization strategies.
EKS platforms from scratch: Helm, namespace-per-tenant isolation, PDBs, NetworkPolicies, External Secrets, Cluster Autoscaler with IRSA.
GitHub Actions pipelines for build, test, scan, and deploy. Automated tenant onboarding with Terraform modules + GitHub Actions.
IAM least-privilege, image scanning, SSL/TLS, policy enforcement, centralized identity management, and secrets rotation.
Built from scratch: namespace-per-tenant isolation, NetworkPolicies, External Secrets, Helm + GitOps. Served multiple development teams across several tenants.
Analyzed node utilization metrics, reduced cluster from 11 to 8 nodes. Implemented Cluster Autoscaler via Terraform with IRSA and designed a rightsizing strategy with Compute Optimizer and Graviton analysis.
Integrated Google Workspace with AWS IAM Identity Center via SAML and SCIM, eliminating manual provisioning. Enforced IAM least-privilege, image scanning, SSL/TLS, and policy enforcement.
End-to-end automation with Terraform modules + GitHub Actions: infrastructure spin-up, secrets provisioning, deploy pipelines, and namespace configuration.
Datadog-based observability: health probes, structured logs, SLO dashboards. Explored open-source alternatives with Prometheus/Thanos, Grafana Loki, Tempo, and OpenTelemetry Collector.
Standardized DevOps practices across multicultural teams without formal authority. Shifted from control-based to enablement-based approach: training sessions, documentation, and tooling that made the right path the easy path.
Looking for a pragmatic DevOps/Cloud partner who can own your platform end-to-end? I bring startup speed with enterprise-grade practices. Let’s talk.