Infrastructure / SRE
Homelab: self-hosted infrastructure as code
75 services in production, DNS HA, full observability and real business automation, all declared as code.
75 services · HA DNS · GFS backup
- Docker Compose
- Traefik
- Prometheus
- Grafana
~ $ whoami
Thales
Machado
Site Reliability Engineer (SRE) · PagBank
availableI run mission-critical production in the financial market: incident response, observability and automation.
At home, the same engineering continues: self-hosted homelab 100% as code, with high availability and full observability.
- SRE · mission-critical
- 7+years
- microservices in production
- 130+
- open source project
- 1
- self-hosted services
- ···
- real-estate crm · saas
- MVP
01about· operating principles
Reliability isn’t improvised, it’s designed.
SRE at PagBank, I sustain the resilience of an environment with 130+ microservices in production, under continuous PCI-DSS audit, with canary pipelines and automation that keep the same problem from happening twice.
I operate infrastructure end to end (network, high availability, observability, security), all declared as code, from design to root-cause analysis.
Outside the bank, the lab is real: a homelab with 75 services in production and wa-courier, the open source WhatsApp gateway I built and maintain. In parallel, I’m building a multi-tenant SaaS for the real-estate market, currently at MVP stage.
Infrastructure as code
Nothing configured by hand that can’t be rebuilt from scratch straight from the repo.
Proactive observability
Metrics, logs and alerts are part of the service design, not an extra bolted on later.
Defense in depth
Multiple layers of defense (network, authentication, log analysis) instead of relying on a single control.
02stack· tools I actually use
cloud & edge
- AWS · ECS · EC2 · ECR · Lambda
- S3 · CloudFront · ELB · WAF
- VPC · Route 53 · API Gateway · ElastiCache
- SQS/SNS · EventBridge · Step Functions
- IAM · KMS · Secrets Manager · SSM
- Cloudflare
observability
- Prometheus · Grafana
- VictoriaMetrics
- Splunk · CloudWatch
platform
- Kubernetes · Docker
- Linux · microservices
- RabbitMQ
iac & automation
- Terraform · Ansible · Salt
- Python · Shell Script
- Go · PHP (básico)
ci/cd & collaboration
- GitHub · GitHub Actions
- Jenkins
- Jira · Confluence
dev & ai
- Node.js · Fastify · Postgres
- Astro · Tailwind CSS
- Claude Code
03projects· case studies
Open Source / BackendWA Courier: self-hosted WhatsApp gateway
Open source HTTP gateway for WhatsApp: one number, one Docker container, zero external service. In production, MIT licensed.
MIT · multi-arch · Prometheus
- Node.js
- Fastify
- Baileys
- Docker
SaaS / Product
Real-estate CRM SaaS
Multi-tenant product in development, from the ground up. Details once the MVP matures.
MVP in progress
- Postgres
- Multi-tenant
- AI
case coming soon
04experience· a decade in infrastructure
PagBank
7+ years- 2025 - now
Senior Site Reliability Engineer
Canary deploy + tuned ASG for critical traffic. The technical gate before production.
- 2022 - 2025
Site Reliability Engineer · Mid-level
Self-service multi-tenant SFTP platform (Jenkins + Terraform). Zero infra queue.
- 2021 - 2022
Site Reliability Engineer
Started as SRE, combining troubleshooting and on-call for payment systems.
- 2019 - 2021
Systems Administrator · TEF
Kept Sitef, AGK and Intellinac running, automating with Salt/Ansible and Terraform.
- 2018 - 2019
Systems Administrator
Ran payment microservices on Docker (Mesos/Marathon), configured via Puppet.
Avançar
5+ years- 2013 - 2018
Support Analyst
Operated and kept the on-prem infra resilient: backup, RAID, firewall, proxy (Squid), VPN and virtualization (VMware), with Shell Script automation and monitoring via Nagios. The base of SRE discipline.
05contact· let’s talk
Open role, project, or infra talk: channel’s open.
Email has the best response SLA. The others are backup.
~ $ contact --channels
email[email protected]
githubgithub.com/thales-machado
linkedinin/thales-machado
locationPassos, Brazil · GMT-3