SR

Sergio Romero Sanchez

Solutions Architect &
Site Reliability Engineer

Twenty years on both sides of the table: scoping and designing systems with the people who pay for them, then running the platforms underneath in production. Based in Costa Rica.

Approach

Seven of those twenty years were customer-facing. As Lead Solutions Architect at IBM I ran on-site requirement sessions with CIOs and technical stakeholders, designed hybrid and multi-cloud target architectures, presented and demoed them to executive and engineering audiences, and stayed with the work through implementation and handoff. At VMware I owned a cloud lab platform end to end — qualifying leads, running discovery calls, delivering live demos, guiding customers to the right subscription — while running the vSphere and NSX infrastructure the platform depended on.

The other side is production. Reliability engineering, SLOs and error budgets, observability, Infrastructure as Code, incident response, and cloud cost optimization across AWS, Azure, GCP, Kubernetes, and OpenShift. The reason both halves matter together: an architecture you have never operated is a guess, and a platform you cannot explain to the person funding it does not get funded.

Outside of delivery work I mentor engineers moving into reliability roles, and I write about what breaks and why. Native Spanish, fluent English.

Track record

Continuous since 2006.

  1. 2024–now

    Senior Site Reliability Engineer

    IBM

    Production operations for a data platform across AWS, GCP, and IBM Cloud. Kubernetes, Terraform, Argo CD, Prometheus, Grafana, SLO and error-budget definition, incident response.

  2. 2019–2024

    Site Reliability Engineer

    VMware

    Owned a cloud training and demo platform end to end: customer discovery and live demos on one side, vSphere and NSX infrastructure on the other. Cut deployment time 25% with Terraform and Ansible.

  3. 2016–2019

    Lead Solutions Architect

    IBM

    Customer-facing architecture engagements. Requirement sessions with CIOs and technical stakeholders, hybrid and multi-cloud design, solution demos, delivery oversight through handoff.

  4. 2013–2016

    Senior Linux System Administrator

    IBM

    Enterprise Linux and virtualization administration across on-premises and cloud environments. Promoted into the architect role.

  5. 2007–2013

    Linux System Administrator

    Hewlett-Packard

    Global enterprise Linux administration, VMware virtualization, and automation scripting for monitoring, patching, and provisioning.

  6. 2006–2007

    Technical Support Specialist

    Sykes

    Bilingual technical support for networking and connectivity issues, in English and Spanish. Where the twenty years start.

Certifications

  • Site Reliability Engineer Profession Expert, Level 2 IBM
  • Certified Kubernetes Administrator CKA The Linux Foundation
  • Certified Solution Architect The Open Group
  • Linux Administrator LPIC-1 Linux Professional Institute
  • Linux+ CompTIA

Projects

Public repositories. Each one is a working lab, not a sample.

  • AWS DevOps GitOps Platform

    Platform engineering on AWS with Terraform, EKS, and Argo CD, including keyless CI/CD through GitHub OIDC.

    Terraform EKS Argo CD

  • SRE Resilience Lab

    Chaos engineering from zero: break pods with LitmusChaos, measure recovery time, verify availability during failure.

    LitmusChaos Kubernetes

  • AWS Cloud Practitioner Hands-On

    Short, focused AWS activities that give CLF-C02 candidates real exposure to core services before the exam.

    AWS Training

  • CKA Study Guide

    Exam preparation built from operational practice: hands-on examples, simulators, and the failure modes that actually come up.

    Kubernetes Certification

  • AWS Instance Scheduler

    Terraform automation for EC2 uptime windows. Start/stop scheduling that takes a visible bite out of monthly spend.

    Terraform FinOps

  • AWS Terraform + Packer

    End-to-end environment provisioning: VPC, subnets, security groups, and an Nginx application behind a load balancer.

    Terraform Packer

Everything on GitHub

Writing

Published in AWS in Plain English.

Everything on Medium

Labs on video

Short, run-it-yourself Kubernetes labs. Every episode has a repository you can clone and break.

  • Chaos on a kind cluster

    Stand up a local kind cluster, inject pod failures, and watch what recovery actually looks like.

    chaos-kind-lab

  • Pod requests and limits

    What throttling and OOMKills look like from the inside, and how to size a workload on evidence rather than habit.

    pod-limits-lab

  • Draining a node without an outage

    Pod disruption budgets, graceful termination, and the drain that takes your service down anyway.

    node-drain-lab

  • Horizontal pod autoscaling

    Drive load at a deployment, watch the HPA react, and find the lag between demand and capacity.

    hpa-autoscaling-lab

  • Burn-rate alerts on an SLO

    Multi-window burn-rate alerting end to end: define the objective, spend the budget, catch it before the quarter does.

    slo-burn-rate-lab

  • Error budgets, explained

    Why an error budget is a spending decision rather than a reliability target, and what changes once a team treats it that way.

The channel @reliabilitybench