I build the platforms, automation, and guardrails that help teams ship reliably and operate with confidence.
My work spans Kubernetes operators, GitOps with FluxCD, secure self-hosted infrastructure, CI/CD, observability, and practical AI-assisted tooling.
I focus on systems that are useful in production: isolated, observable, repeatable, and documented.
I verify live, then write it down.
Based in Portugal · Website · LinkedIn
A Go operator for fraud-scoring stream pipelines, with consumer-lag scaling, SLO-based self-healing, automated canary rollback, and FluxCD as the source of truth. The project is backed by live runs, verified timings, FAQs, and postmortems.
A design study and Ansible template for isolated self-hosting with Incus, rootless Docker, ZFS, and per-tenant Cloudflare Tunnels. Built around a simple principle: each tenant should be able to operate without discovering its neighbours.
- Build platforms people can trust. I make systems observable, explainable, and calm under failure, so teams can ship without guessing what breaks next.
- Verify live, then document the result. Features are tested against real failure modes, not just successful local runs.
- Treat GitOps as an operating discipline. Git remains the source of truth, and platform behavior stays explicit and reviewable.
- Lead with practical solutions. I care about clear trade-offs, useful automation, and giving engineers the context to move independently.
- AI-assisted, engineer-driven. I use AI agents to accelerate exploration and delivery; I set the direction, review the work, and stay accountable for the outcome.
I'm open to conversations about platform engineering, infrastructure, technical leadership, and reliable product delivery.
