A showcase of my journey building resilient infrastructure and automating workflows.
π‘οΈ Ez-Lab: Cloud-Native Edge Infrastructure
Infrastructure as Code / DevSecOps / GitOps
Architected and provisioned a resilient, bare-metal microservices environment on a Raspberry Pi 5. This project operates as a production edge cluster protected by Zero Trust principles and automated delivery pipelines.
- Architecture & Orchestration: Docker Compose managing the *Arr Stack, internal APIs, and Pi-hole for network-wide DNS filtering.
- CI/CD & GitOps: Automated a zero-downtime software delivery pipeline utilizing GitHub Actions and Watchtower to build, push to a private registry (GHCR), and deploy cross-platform containers.
- Zero Trust Security: Secured internal network traffic and bypassed ISP CGNAT by deploying Cloudflare Tunnels and Tailscale mesh VPNs, successfully eliminating public-facing open ports.
- Observability & Chaos Engineering: Deployed Prometheus and Grafana for real-time cluster telemetry and visualization, alongside custom Chaos Agents simulating container failures to validate auto-recovery policies.
- Status:
Production π’
πΈ Live Infrastructure Status (Portainer)

Tech Stack:
DockerGitHub ActionsCloudflare Zero TrustPythonLinuxBashPrometheusGrafana
βοΈ MyssTic Warden: Enterprise AWS Architecture
Infrastructure as Code / Event-Driven / Zero Trust
Designed and provisioned a highly available, event-driven AWS architecture from scratch using Terraform, focusing on immutable infrastructure, configuration management, and automated incident response.
- IaC & Remote State: Fully managed via Terraform with state locked in DynamoDB and encrypted in S3 (AES-256).
- Shift-Left Security & IAM: Enforced security compliance via Checkov and Trufflehog directly in the pipeline. Replaced all static deployment credentials with secure AWS OIDC (OpenID Connect) federation.
- Configuration Management: Standardized base setups and system packages securely across AWS instances utilizing automated Ansible playbooks.
- Zero Trust Networking: Custom VPC architecture isolating public DMZs from private database subnets, strictly controlled via Security Groups.
- Database Isolation: Multi-AZ PostgreSQL deployment on Amazon RDS, integrated with AWS Secrets Manager for dynamic, zero-touch credential injection.
- Serverless Chaos Response: Implemented an event-driven observability pipeline. CloudWatch monitors EC2 CPU anomalies, triggering an SNS topic that invokes a Python AWS Lambda function to push critical alerts directly to a Telegram Bot.
- Status:
Production π’
Tech Stack:
TerraformAWS (EC2, VPC, RDS, IAM)Serverless (Lambda, SNS)PythonCloudWatch
πΊοΈ The SRE Roadmap (Engineering Log)
Continuous Learning / Documentation
I am publicly documenting my transition to Site Reliability Engineering through a structured 11-Phase Roadmap.
- Phase 1-9.5 (Completed): Infrastructure Hardening, Zero-Trust Architecture, Shift-Left Security workflows, Ansible Automation, and Multi-Arch CI/CD pipelines.
- Phase 10 (Current Focus): Advanced Cloud Security, Hardening, FinOps optimization, and Hybrid Edge SOC development.
- Phase 11 (The Final Boss): Kubernetes Orchestration & GitOps with ArgoCD.
π Knowledge Base Modernization (Konecta)
Docs-as-Code / Process Optimization
Led an initiative to modernize internal technical documentation for Tier 2 Support agents.
- Problem: Critical troubleshooting info was scattered across legacy static sites and PDFs.
- Solution: Migrated content to a Docs-as-Code system using MkDocs, enabling version control and faster search.
Tech Stack:
MkDocsMarkdownGit
β Before: Legacy Static Site

β After: Docs-as-Code (MkDocs)

π SRE Portfolio & Live Telemetry
Hybrid Cloud / GitOps / Multi-Architecture CI/CD
More than just a static website, this portfolio is a live demonstration of Enterprise Cloud Architecture. The frontend fetches real-time telemetry from a Golang API, deployed across a dual-environment infrastructure: a local Raspberry Pi Edge cluster (Staging) and an AWS Graviton EC2 instance (Production), entirely managed through a Zero-Intervention GitOps pipeline.
- Frontend: Hugo + Nginx (Alpine Slim)
- Backend: Golang 1.26 REST API (Distroless Container)
- Infrastructure: AWS (ECR, Route53, VPC) + Raspberry Pi (Edge)
- Security & Routing: Traefik Ingress + Let’s Encrypt (Prod) / Cloudflare Zero-Trust (Staging).
ποΈ Live Architecture
graph TD
classDef dev fill:#1a1b26,stroke:#7aa2f7,stroke-width:2px,color:#c0caf5
classDef cicd fill:#24283b,stroke:#bb9af7,stroke-width:2px,color:#c0caf5
classDef edge fill:#152515,stroke:#44aa44,stroke-width:2px,color:#fff
classDef cloud fill:#221535,stroke:#ff9900,stroke-width:2px,color:#fff
classDef net fill:#1f2335,stroke:#7aa2f7,stroke-width:1px,color:#a9b1d6
%% FLUJO DE CΓDIGO
Developer((π» Ezequiel)):::dev -->|git push| GitHub{π GitHub Actions Engine}:::cicd
subgraph "CI/CD DevSecOps Gates"
GitHub -->|1. Audit| Trufflehog[π΅οΈββοΈ Trufflehog Secrets]:::cicd
GitHub -->|2. Lint| Hadolint[π³ Hadolint Docker]:::cicd
GitHub -->|3. Scan| Trivy[π‘οΈ Trivy Vulnerabilities]:::cicd
end
%% ENRUTAMIENTO POR RAMAS (LOGICA DEL PIPELINE)
Trivy -->|Branch: develop| GHCR[(π¦ GitHub Registry)]:::cicd
Trivy -->|Branch: main| ECR[(π³ AWS ECR Private)]:::cloud
%% ENTORNO STAGING (RASPBERRY PI)
subgraph "Ez-Lab Environment (Staging Edge)"
GHCR -.->|Auto Pull| WatchtowerPi[π Watchtower Agent]:::edge
WatchtowerPi -->|Deploy :stage| WebStage["π₯οΈ Frontend (Hugo)"]:::edge
WatchtowerPi -->|Deploy :stage| APIStage["βοΈ Telemetry (Go)"]:::edge
WebStage -.->|CORS Fetch| APIStage
end
%% ENTORNO PRODUCCIΓN (AWS)
subgraph "MyssTic Warden Environment (Production Cloud)"
ECR -.->|IAM Auth Pull| WatchtowerAWS[π Watchtower Agent]:::cloud
WatchtowerAWS -->|Deploy :prod| WebProd["π₯οΈ Frontend (Nginx)"]:::cloud
WatchtowerAWS -->|Deploy :prod| APIProd["βοΈ Telemetry (Go)"]:::cloud
Route53((π Route 53)):::cloud --> Traefik[π Traefik Ingress]:::cloud
Traefik -->|Path: /| WebProd
Traefik -->|Path: /api/health| APIProd
WebProd -.->|Internal VPC Fetch| APIProd
end
- $ Establishing Zero-Trust handshake...
Tech Stack:
GolangHugoNginxAWSECRGHCRGitOps