320K–485K USD / year
Principal Engineer, Capacity/Compute
DevOpsHybrid — San Francisco, CA
Published on 2026-11-20
These details were extracted automatically from the original listing. They may not be complete or fully up to date — it's worth checking the original job post.
About this role
Build production systems that help Anthropic account for, plan, and efficiently allocate its large, multi-cloud compute fleet. The role spans telemetry pipelines, Kubernetes operations, observability, workload efficiency, cost attribution, and forecasting.
About the company
Anthropic builds reliable and interpretable AI systems, including the Claude family of models, for businesses and developers. The company focuses on AI safety research and practical AI assistants for enterprise and consumer use.
The team
Capacity Engineering
Stack
PythonSQLBigQueryKubernetesPrometheusPromQLGrafanaAWS
What you'll do
- Build capacity planning and allocation tools, including placement, queueing, guardrails, and occupancy KPIs.
- Improve fleet and workload efficiency through utilization measurement, benchmarking, rightsizing, and unused-capacity recovery.
- Reconcile provider billing, telemetry, and internal systems to attribute costs and forecast compute demand.
- Develop data pipelines into BigQuery with ownership of completeness, latency SLOs, and gap detection.
- Operate Kubernetes-native collection, labeling, reservation, and scheduling systems.
- Define data contracts and build reliable, discoverable products for engineering, finance, and leadership consumers.
What we're looking for
- Track record building and operating production systems in hands-on engineering or DevOps roles.
- Production-quality Python and SQL skills.
- Deep operational experience with at least one major cloud provider.
- Experience with Prometheus, PromQL, Grafana, recording rules, and production monitoring.
- Able to gather requirements and collaborate across teams in ambiguous environments.
Nice to have
- Capacity planning, resource management, or cost attribution at hyperscaler or large-scale ML infrastructure.
- Scheduling and packing efficiency or profiling distributed workloads.
- Multi-cloud ingestion and normalization of billing, reservation, and telemetry data.
- Infrastructure cost forecasting and total-cost-of-ownership analysis.
- Accelerator metrics or hardware-level ML training and inference experience.
- Internal data products, schema contracts, self-service access, or exabyte-scale storage lifecycle programs.
Benefits
- Flexible working hours
- Generous vacation and parental leave
- Optional equity donation matching
- Hybrid policy: staff are expected to work from an office at least 25% of the time
