Site Reliability Engineer, Compute
Fluidstack · San Francisco, CA
About this role
## About Fluidstack Fluidstack exists to make humanity more free. We’re building civilization-scale infrastructure for AI—acquiring power, designing and operating data centers, and delivering **10 to 100s of GWs of compute faster** than anyone else. ## How We Operate - **Be a barrel:** full autonomy, own things end-to-end, no permission required to operate outside your core role. - **Insane urgency:** drive everything forward as fast as possible. - **Reason from first principles:** challenge assumptions; the best idea wins. - **Love of the game:** long hours at high intensity to push the frontier forward. - **Build something that actually matters:** spend your time on work with real-world impact. ## Production Engineering Team — What You’ll Work On - **Repair pipeline at fleet scale:** automate chip fault detection → triage → RMA → return to service (no human intervention). - **GPU qualification in ~6 months:** burn-in, performance baselining, and NPI execution to define “production-ready” before sites go live. - **Migrate live compute at construction speed:** convert clusters across production sites; make Kubernetes-orchestrated bare metal sustainable at multiple GW/year. - **Real-time fleet observability & orchestration:** build the layer that makes hyperscale AI compute operable; debug, tune, and performance-test infrastructure as the fleet grows. ## Role Scope (Site Reliability Engineer, Compute) - **Own compute fleet health end-to-end:** build metrics pipelines, alerting, and a unified health view across **Kubernetes-orchestrated workloads and bare metal**. - **Turn deployment/repair into a pipeline:** automate compute failures from detection through triage, parts management, and return to service—no one-off scripts or heroics. - **Design and expand the GPU qualification platform:** burn-in, performance baselining, and NPI execution; define what “good” means before production. - **Own Redfish and BMC tooling:** firmware-level telemetry, fleet-scale log collection, and low-level access layers used by repair and health tooling. - **Reliability, scalability, and operations at scale:** aggressive automation, tooling, and incident discipline to operate one of the largest GPU fleets in the world. ## What We’re Looking For - Treat **toil as a bug**; manual steps belong on a backlog. - Strong **hardware instincts**; comfortable reasoning about failure modes at firmware/silicon levels. - Move toward **ambiguity**; build the map and explain it. - Learn quickly; reach real competence in unfamiliar domains. - **Pager-ready:** run incidents, write postmortems, fix systemic causes, and move on. - Fluent with **AI tooling** (LLM APIs, MCP servers, agentic frameworks); use tools like Claude Code/Cursor daily. - Shipped **production automation** other teams depend on; comfortable using AI coding tools in any language. - **Bonus:** hardware lifecycle management & RMA automation; BMC/Redfish/IPMI tooling; GPU qualification/burn-in frameworks; workflow/orchestration engines (Temporal, Cadence); metrics/alerting (Prometheus, Grafana); Go or Python. ## Salary & Benefits - Competitive total compensation (**salary + equity**) - Retirement/pension plan (local norms) - Health, dental, and vision insurance - Generous PTO (local norms) - Fluidstack is committed to pay equity and transparency ## Equal Employment Opportunity Fluidstack is an Equal Employment Opportunity Employer. ## Apply https://jobs.ashbyhq.com/fluidstack/05c2e69c-42f9-4fcb-9cf0-a467aa
Listing freshness
CronJobs last confirmed this listing 1h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.