Zbigniew Zabost

Cloud SRE team behind HiveMQ Cloud, a managed MQTT service on multi-cloud Kubernetes (30+ clusters on EKS, AKS, GKE). Led a fleet-wide upgrade with a GitOps merge train I designed (health gates; LLM judge only for edge cases). Drove fleet standardisation and the Tailscale operator rollout. Root causes turned into guard rails (JVM memory, K8s DNS, cert-manager). Claude Code skills for ops; AI triage agent that won an internal hackathon. Enterprise 2.0 PoC (Go, Temporal). 24/7 on-call.

First SRE role: HiveMQ Cloud's dedicated enterprise deployments on AWS, Azure and GCP. Helped build the first GKE cluster and GCP enterprise deployment, then delivered more on my own. Co-owner of the Helm chart behind all dedicated deployments; first Helm unit tests. Rolled out a unified chart on 3 clouds and migrated a customer onto it. SIEM integration across AWS, Azure, GCP. Led a large customer migration (reverse-engineered an undocumented setup). 24/7 on-call.

Platform team. Built and migrated the company-wide CI/CD platform as a one-person team: legacy on-prem Jenkins to bare-metal cloud servers (Equinix Metal, ESXi; Jenkins, Nexus, Gradle Enterprise, Grafana) before a hard deadline. Tested recovery before go-live: when OpenStack + Ceph failed a simulated controller outage, pivoted to a stack I could operate. Automated with Terraform and Ansible, documented and handed over.

Instana was acquired by IBM; employment transferred to IBM in Jan 2022, same role. Engineering and delivery enablement: one-click OpenShift cluster provisioning (a Jenkins job, minutes instead of a semi-manual process); release-test pipelines for the Instana agent across OpenShift/Kubernetes versions and install methods; IaC with Terraform, Packer, GCP; CI with Jenkins and Concourse; backend tests in Go.

CI/CD and infrastructure tooling for Groupon's global engineering. Delivered a cloud CI on AWS (Jenkins, Docker, Terraform) and drove the migration off Mesos/DC/OS, with automated repo migration across 700+ microservices; up to 1,100 builds/day, became the standard CI. IaC CI/CD tooling used by 15 teams. Jenkins plugins measuring commit time to staging (CTTS). Docker image retention saving TBs. PagerDuty rotation; workshops in 4 cities.

Led the project introducing and automating HAProxy Enterprise for the API services, configured through Chef. Introduced Chef development standards and a CI pipeline for the infra team's cookbooks (Test Kitchen, ChefSpec, InSpec); before that, cookbooks weren't tested. Promoted to Senior about a year after moving into systems engineering.

Moved internally from test engineering into systems engineering. Ran and extended the ELK and Graphite monitoring stacks and the delivery tooling (Jenkins, Nexus, Docker Registry); built internal tools in Java and Python; on-call. Organised workshops and tech lunches.

Joined as zanox; the company was later renamed Awin. Introduced automatic production deployments, starting with one greenfield project; it became the team standard for new applications. Steered the team towards operating its own applications. Test automation, performance testing (JMeter, Gatling), monitoring (ELK, Splunk, Graphite).

Test automation for Groupon's web platform; after a reorganisation, also tech lead of a small QA team doing integration testing across the platform. Introduced Selenium Grid, a multi-locale Selenium suite and a JBehave-based testing DSL; worked with release engineers on the build and release pipeline.

Started in Łódź as a developer on a Java web application; from April 2011 on site at Nokia (gate5) in Berlin, in an international team focused on test automation, CI and user acceptance testing.