Expert: Interactive Coding, Site Reliability Engineering (Contract, Remote)

Role overview Work through production-style reliability and infrastructure problems in a sandboxed environment, using the judgment expected during real operations. In live screen-share sessions, you’ll investigate symptoms, mitigate or fix issues, explain your reasoning, and point out where a scenario does not reflect operational reality. The resulting work helps define reliable AI performance on site reliability engineering tasks; prior AI experience is not required.

Responsibilities - Triage operational alerts and investigate system behavior using metrics, logs, and traces. - Diagnose root causes and choose appropriate mitigations, rollbacks, or fixes. - Work with infrastructure and delivery tooling such as Kubernetes, Prometheus, Grafana, Helm, Terraform, and CI/CD systems. - Explain decisions during shared-screen sessions and identify unsafe or misleading approaches. - Write a brief account of the investigation and resolution after each scenario.

Requirements - Professional experience in site reliability, production engineering, platform operations, or infrastructure. - Practical experience with the systems and tools relevant to your specialty, such as orchestration, monitoring, infrastructure as code, or deployment pipelines. - Sound incident judgment: ability to investigate evidence before changing configuration and to weigh the risks of potential fixes. - Clear verbal reasoning during live work and concise written communication afterward. - Availability for live sessions with a few hours of overlap with US or European working hours.

Nice to have - Experience on a production on-call rotation, including incident response and post-incident reviews. - Experience writing runbooks, post-mortems, service-level objectives, or incident training materials. - Familiarity with failure modes of AI coding agents in operations work.

Benefits and work setup - Independent contractor engagement, fully remote and open internationally. - Approximately 10 hours per week for an initial one to three months, with possible extension. - Pay is $150 per hour; the trial session is paid. - Selection includes a short interview and resume review, followed by a paid trial scored against a quality checklist. - Work must be original and human-authored. Use fictional scenarios and do not share confidential configurations, runbooks, incident details, or other protected material.