Site Reliability Engineer resume example

• Built observability for Kubernetes services with Grafana burn-rate dashboards and on-call runbooks; tied each page to an SLO and a tested mitigation rather than a raw CPU threshold.
• Rehearsed Linux node replacement from reviewed Terraform plans, checking workload drain, storage mounts, and recovery time; recorded a mount dependency that had been missing from the recovery procedure.
• Built a Go probe for the customer-facing request path and a Python recovery check for restored records. Used both in a failover exercise to detect healthy processes serving incomplete data; updated the runbook and verified the correction in a repeat drill.
• Reconstructed an incident timeline from traces, deploy events and SLO alerts; converted the triggering dependency failure into a recovery drill with measured restoration time.
• Owned an on-call recovery drill using Prometheus alerts and Grafana traces, resolved a Kubernetes dependency failure and completed rollback within the agreed SLO window.
• Built Python debugging checks using Linux logs and Terraform state, reduced incident triage from 32 to 14 minutes and released tested CI/CD recovery instructions.

• Led an incident response review using request traces and a timestamped timeline, assigning owners to retry limits and queue backpressure fixes. Closed the failure path with a replay test rather than only updating the runbook.
• Reviewed a proposed change against the existing interface contract, reproduced a reviewer’s edge case, and added the missing test or example before merging the change.
• Investigated a failure reported by a downstream team using the exact version and input they supplied. Reduced it to a reproducible test, delivered the fix with compatibility notes, and confirmed the original case passed before closing the report.

• Replaced manual production changes with reviewed Terraform plans, drift checks, and staged rollout gates
• Built an isolated fixture with normal, malformed and interrupted inputs; recorded the expected response before running the implementation so failures were distinguishable from a changed assumption.
• Packaged the reproduction steps, configuration and regression tests with the project; checked the instructions from a clean environment and documented the unsupported cases instead of claiming production readiness.
• Owned the independent regression fixture using isolated inputs, resolved a failing edge case and released a reproducible test package with expected outputs.

Relevant coursework: Algorithms, operating systems, databases, computer networks

• Published an independent technical note using the project’s before-and-after traces, explaining the failed assumption, regression coverage and cases the implementation does not support.