Pedro Lucas

I’m a DevOps / Site Reliability Engineer with hands-on experience owning cloud production infrastructure end to end—GCP and AWS—plus Kubernetes-based operations using Kubernetes/K3s. I build and run CI/CD and incident-ready systems that keep high-traffic e-commerce and computer-vision platforms stable.

One of my strengths is observability: I’ve built observability stacks from scratch using OpenTelemetry, VictoriaMetrics, Jaeger, and Grafana, and added automated alerting. I also led root-cause investigations to resolve production-down database and API failures, improving time-to-detect for incidents.

I’m comfortable driving reliability and performance across distributed systems—tuning MongoDB clusters, tracing concurrency spikes, and fixing production issues such as Bearer Token caching bugs. In parallel, I deliver measurable cost impact through FinOps re-architecture, including cutting cloud security spend by 84% (~R$8,700/month) and reducing GCP Cloud Armor spend by right-sizing and offloading edge filtering to Cloudflare Geoblocking.

In computer vision work, I’ve architected production pipelines and services (NATS JetStream, MongoDB, Redis, FastAPI, Nginx), deployed and managed a K3s cluster, and integrated monitoring with Grafana, Prometheus, and the ELK Stack. I enjoy collaborating in distributed, English-speaking teams (B2) and using automation (Terraform, Ansible, GitOps) to make operations repeatable.