Site Reliability Engineer

More than 9,000 organisations use our HR and payroll suite, from regional hospitals to school districts. Annual releases gave way to two-week sprints years ago, and the platform now handles over 200 million API calls a day. Founded in 2008, we now have 620 employees. On-call rotates weekly and pays a flat stipend whether or not the pager goes off.

Job summary

We're looking for a Site Reliability Engineer to join our team. This is a remote role. Our services run on Kubernetes across two cloud regions. This hire will help us get better at capacity planning and failover. SRE role with a heavy software component. We expect you to write code that removes toil, not just run playbooks. You can work from home anywhere in Hungary. We have no office requirement. Normal working hours, Monday to Friday. We use Slack, Notion and Google Meet. Cameras on for team calls.

Essential duties and responsibilities

Review architecture proposals for reliability risks before launch. Run capacity planning and load tests with k6, Locust or Gatling. Improve observability with Prometheus, Grafana, Loki, Tempo or Datadog. Harden databases and caches with replication, backups and restore drills. Maintain on-call schedules, runbooks and escalation policies in PagerDuty or Opsgenie. Lead incident response as incident commander and run blameless postmortems. Plan and run game days and chaos experiments. Write tooling in Go or Python to automate failover, scaling and routine operations.

Required qualifications

3+ years in SRE, production engineering or infrastructure roles. Deep Linux knowledge, including systemd, networking and tools like strace and tcpdump. Clear writing for postmortems and runbooks. Understanding of distributed systems: consensus, replication, partitioning, backpressure.

Preferred

Multi-region or active-active architecture. EBPF-based tooling.

What we offer

Private medical care. Extra annual leave days. Language courses. Meal vouchers.

Start: Flexible.

Demo job created on 11jobs for testing and demonstrations. It is not an actual opening at DCD Data Bt..