Site Reliability Engineer · 20+ Years Building Resilient Systems
Rafael Dihl
Reliability-focused engineer specializing in distributed systems, observability, and incident response — keeping high-QPS, member-facing services resilient at scale.
About
20+ years engineering and operating large-scale systems, with 7+ years focused on the reliability, observability, and automation that keep production services healthy. I take a data-driven approach to reliability — defining SLOs, instrumenting systems with Datadog and OpenTelemetry, leading incident response and post-incident analysis, and designing for failure modes across distributed systems. Currently at Fanatics, running high-availability workloads on AWS EKS with GitOps delivery (ArgoCD) and automated, right-sized infrastructure. Previously at Meta (via Magnit), where I designed and deployed the AWS environment behind the Wolfram Alpha LLM integration, engineered for high QPS, low latency, and strict privacy compliance. Comfortable across Python, Go, and Java, and fluent in both English and Portuguese.
Impact
Resilience & disaster recovery. Architected and deployed multiple enterprise HashiCorp Vault clusters on-premises with disaster recovery — configurations that directly enabled recovery from several high-visibility production outages, minimizing member-facing downtime.
Observability & incident response. Built infrastructure monitoring with Datadog and Terraform — custom dashboards and alerting rules that reduced incident response time — and led the FluxCD→ArgoCD migration to gain deployment traceability and fast, reliable rollbacks on AWS EKS.
Reliability at scale. Provisioned and deployed the AWS environment (Terraform) to run the Wolfram Alpha LLM as an internal service, engineered for high QPS, low latency, and strict privacy compliance across Meta traffic.
Reliability vs. cost. Applied a data-driven capacity analysis to a batch service, identified unused elastic capacity, and safely released it — reducing server usage by 60% and translating to $7M in annual savings with no loss of reliability.
Right-sizing & efficiency. Drove the migration of EKS workloads to AWS Graviton (arm64) and introduced Karpenter for dynamic, demand-driven node provisioning. Analysis of the 233-host fleet identified 147 eligible x86 instances, projecting ~$83K in annual savings — scaling to ~$416K/year at 5× fleet growth. Karpenter provisions nodes right-sized to actual pod requirements and automatically consolidates or terminates underutilized capacity, where fixed Managed Node Group pools would leave idle nodes running.
Reliability as a standard. Designed and delivered CI/CD pipeline libraries that standardized safe, repeatable delivery across the entire organization — spanning Kubernetes, Helm, Terraform, Vault, Datadog, and more.
Experience
Senior Platform Engineer
- Architected and maintained scalable CI/CD pipelines on AWS EKS using GitHub Actions, FluxCD, and ArgoCD, including provisioning ArgoCD clusters from scratch.
- Led the migration from FluxCD to ArgoCD to improve deployment traceability, centralized visibility, and rollback capabilities.
- Built and managed infrastructure monitoring with Datadog and Terraform, creating custom dashboards and alerting rules to reduce incident response time.
Software Engineer — Contingent Worker @ Meta
- Designed and implemented AWS infrastructure (EC2, EKS, RDS, Kafka, Terraform, Ansible, Helm) to deploy the Wolfram Alpha LLM solution for Meta traffic, ensuring high QPS, low latency, and privacy compliance.
- Spearheaded enhancements to data and code delivery pipelines, significantly improving build efficiency and deployment velocity using Kubernetes, Golang, Python, PHP, Bash, and Terraform.
Senior DevOps Engineer — Consultant @ Charter Communications
- Developed and delivered Infrastructure as Code solutions enabling partner engineering teams to migrate applications to AWS.
- Designed CI/CD pipeline libraries that standardized the use of GitLab, Kubernetes, Helm, Docker, Terraform, HashiCorp Vault, Fluentbit, Datadog, and Splunk across the organization.
Senior DevOps Engineer — Consultant @ DealerTrack
- Provisioned and deployed multiple enterprise HashiCorp Vault clusters on-premises with disaster recovery; configurations played a key role in recovering from several high-visibility outages.
- Led development of a scalable CI pipeline (Jenkins CasC + OctopusDeploy) stabilizing builds and deployments across 10+ environments.
- Containerized a large legacy monolith into a 12-factor app, achieving 40% performance improvement through Docker and immutable infrastructure.
Software Development Advisor
- Designed and implemented large-scale Java EE and SOA applications in globally distributed teams, acting as Subject Matter Expert in field services.
- Coordinated work across small to medium teams on strategic, tactical, and maintenance projects; provided L3 production support.
Senior Software Developer
- Developer on ADP Portal R8, one of ADP's most widely used products with 17M+ users.
- Built features using Struts, WebSphere Portal, Application Servers, JSF, and JMS.
Skills
Reliability
Languages
Infrastructure
CI/CD & GitOps
Data & Messaging
Observability
Platform & Security
Education & Certifications
Pontifícia Universidade Católica do Rio Grande do Sul
Bachelor of Information Systems (BCompSc) · Porto Alegre, Brasil
Oracle Certified Associate — Java SE 7 Programmer
Oracle · Credential ID: OC1304424
Certified Scrum Master
Scrum Alliance · Credential ID: 000447682
Beyond Work
Running
Completed 3 marathons (Chicago, London, Porto Alegre) and currently training for Berlin in September 2026.
Reading
Avid reader who enjoys exploring new ideas and perspectives through books.
Beach Time
Love spending time at the beach with my family, enjoying the Florida coast.