Build the software and automation behind that path: hardware telemetry and alerting (BMC and Redfish exporters, node-level metrics, log collection), removal of manual steps from diagnostics, production drain and return, alert silencing, and post-repair validation, and component-level tracking from detection to recovery. Own the end-to-end hardware repair lifecycle for the Starlink server fleet and its components (CPUs, memory, SSDs, NICs, optics, and power supplies): triage alerts, isolate failures to the component level, and drive safe removal from production, repair, validation, and return to service.